tatvic, Author at Tatvic Analytics

conversion optimization services product conversion rate ecommerce conversion analysis

Posted on November 28, 2021 - 6:50 am by tatvic

This post aims to develop an understanding of how to calculate product-specific conversion rates and discuss the potential analysis & insights that can be generated from it.

Quite often the business problems or questions that most e-commerce websites encounter are one of the following:

How much product to stock?
Which products are fast-moving?
How can I move some of the products quickly?
How to measure how many views different products have received?
Are the product views turning into transactions & revenue? What actions can I take based on such data?

What is the product conversion rate?

There are many ways to define what we can call the product conversion rate. For example, you can check the number of transactions for any given product divided by the number of visits to that specific product. Moreover, you can also use the quantity to no. of visits ratio to calculate the product conversion rate.

Based on the product conversion rate, an automobile e-commerce website was able to update its recommendations section, which in turn led to a 20% increase in transactions.

For our understanding, we will use the ratio of quantity sold to the no. of visits for the different products as a product conversion rate.

Why product conversion rate?

Based on the product conversion rate, an automobile e-commerce website was able to update its recommendations section, which in turn led to a 20% increase in transactions.

For our understanding, we will use the ratio of quantity sold to the no. of visits for the different products as a product conversion rate.

How to get data from Google Analytics?

There are two approaches that can be used for obtaining the data for product views:

Approach 1: Ecommerce / Enhanced Ecommerce:-

Here is the syntax that you can use to generate product views statistic in event tracking:

Event Category: Enhanced Ecommerce
Event Action: Product Detail View
Event Label: {{current page path}}

This needs to fire on all your product pages in Google Analytics.

As a result of this, you will be able to create a custom report with “Product” as the dimension, and “Product Detail Views” as the metric.

Report Image Source: Google Merchandise Store

Approach 2: Custom Events Implementation:-

You can implement the following data layer snippet to pass an event every time a product detail page is loaded.

Event Category: Product Detail View
Event Action: {{Product Name}}
Event Label: {{current page path}}

Based on this implementation, you will get the Product Names in your event tracking report. You will not need to go through the process of implementing E-Commerce tracking on your website, and will still be able to get unique product views as well as total product views of each product.

By putting this data together, here is what you can get as an output for different products. Please note that in the third column you can create conversion rates for your different products.

Product Name (based on Event Action)	Unique Views (based on Unique Events)	Quantity Sold	Product Conversion Rate	Product Revenue
Google Utility BackPack	260	9	3.46%	2231
Chrome Dino Collectible Figurines	658	29	4.41%	1840
Google Black Eco Zip Hoodie	713	27	3.79%	1721
Google Incognito Zip Pack	322	35	10.87%	1611
Google Eco Tee Black	133	15	11.28%	1364
Google Incognito Techpack V2	413	15	3.63%	1160
Google Campus Unisex Zip Hoodie	142	17	11.97%	1119
Chrome Dino Marine Layer Tee	300	12	4.00%	1079
Noogler Android Figure	250	20	8.00%	1872
Google Cloud Unisex Zip Hoodie	129	24	18.60%	1654

Some of the products have a high conversion rate as shown in highlighted portion. These are your winners.

Plot this conversion data with the product’s revenue and you will be able to analyze various aspects of your product performance indicators, including:

Know whether high converting products are getting sold or not
Use the advanced segments to calculate the product conversion rate for different traffic sources
Figure out low-converting products, reduce them from your stocks and save money from the warehouse!

Ecommerce Predictive Analytics Transformation Guide

Posted on November 18, 2021 - 6:30 am by tatvic

When the COVID-19 pandemic broke out, much of the world moved online, accelerating a digital transformation that has been underway for decades. From a buzzword to a current-day reality, e-commerce in India has been experiencing remarkable growth, successfully changing the way people transact. People today can shop literally everywhere within minutes, be it their workstations or homes, and most importantly, at any time of the day at their leisure.

eCommerce as an industry is taking a huge evolutionary leap. We’ve come to a point where digitally native companies are working around voice commerce, partnering with local businesses on exclusive ads, and most of all testing a mix of Virtual and Augmented Reality giving impactful experiences.

How can eCommerce Analytics help here?

Considering the current state of the eCommerce Industry, Analytics can help in keeping an eye on the value generated by business and business growth. There are tremendous ways through which eCommerce Analytics can help in increasing the user base by targeting the specific audience to increase the conversion rate like:

Marketing analytics
Segmentation Analysis
Predictive Analytics

What are the capabilities of Predictive Analytics?

Predictive Analytics provides e-commerce businesses with a deeper understanding of customer habits and preferences. When coupled with advanced Machine Learning capabilities, it can correlate data from different sources to create personalized recommendations for particular customers or segments.

What problems can be addressed through Predictive Analytics?

How to give each customer the best shopping experience possible
What products are in demand currently- and are likely to be in demand in future
How to Create Customized Marketing Messages
Who is the right set of people to remarket to?
How much to spend on remarketing channels?
What would be the right medium to invest in?
Users who are likely to buy a particular product
Users who are likely to purchase a product twice a month
Users who are likely to use credit as a mode of payment
Users who are likely to add the product to wishlist

How can Tatvic help in Predictive Analytics?

PredictN is a machine learning model that generates an Audience Segment (Qualified Visitors) who carry the probability of getting converted as a customer within the next ‘n’ days.

How does it work?

For more details visit here

Why should the eCommerce industry do Predictive analysis?

The internet, custom-made Apps, and Mobiles have undoubtedly changed the way we buy whether it is food, clothing, automobiles, etc. Due to the ease of availability options at finger tips, people have moved from Retail to Online, as it offers more convenience and offers.

To understand the data and make business-centric decisions, many companies are looking forward to cross-sell and up-sell their products and services to their customers by analyzing the data and targeting specific sets of audiences for remarketing. eCommerce analysis will help the companies gain an edge over their competitors, as they will know exactly whom to target to get maximum conversions, and what are the trends/products/services likely to create a boom in the future. Below mentioned are some of the techniques common techniques that companies use:

Social Selling

Social commerce offers a much more engaging experience, as the majority of the people, are socially active on the various social media platforms. People love the way in which products are displayed with short videos to highlight the central message. This way of social nudge helps drive instant sales.

Dynamic Pricing for Optimal Sales and Profit

To stand out in this highly-competitive environment, eCommerce companies need to monitor their competitors’ prices, analyze seasonal and historical demand, and react to These insights in nearly real-time. With the historical data of supply and demand, prediction can be more accurate with the needs of essential products geographically.

Subscription-Based Business Models

Subscription-based eCommerce is booming and has recently gained a lot of traction with the ease of automated direct debit on the monthly basis and many more. The subscription model holds the power to lure in new customers through Word of Mouth and have high conversion rates.

Real-time conversation using Chatbot

Consumers have multiple queries while buying the product online and support executives may not be available all the time. With the help of AI Chatbots, the problem can be resolved. Chatbots are widely used by many eCommerce companies to provide real-time support for customers.

Conclusion

The future of eCommerce is not just limited to this, there is a way more to come. But summarizing the bit, analytics would also be evolving and finding new capabilities to generate insights on the same.

5 Google Analytics Reports that are Helping eCommerce Merchants Increase Conversions

Posted on November 2, 2021 - 9:48 am by tatvic

For eCommerce business, the most important business objective is to increase the conversion/User spends and ROI of the business. To do that it is of utmost importance that you analyse your data and make data-driven decisions because you just don’t want to waste your hard-earned traffic on a plethora of multivariate tests that just won’t contribute to increasing your conversions.
After all, you’ve spent a lot of time and resources growing your business.

But the question might arise, How do I get data and analyse it to make business decisions?

Considering you are already a Google Analytics user, It comes with a lot of default reports that you can use to analyse and make data-driven decisions.

If you are running an eCommerce business, then a big congratulations!

Because Google Analytics comes up with a dedicated section designed for eCommerce clients which provides you with all kinds of eCommerce reports that you have ever thought of. It includes conversion funnels starting from user acquisition to conversion action. Using this you can analyse all the data points which you might ever need. Check those reports under the Conversions → Ecommerce section.

This analysis can lead you to pinpoint areas that are not working too well. Going deep inside will reveal the most likely problems. This will narrow down the changes and improvements you can make with your eCommerce optimisation strategy.

Sounds easy? It is never that easy.

Don’t worry! We are here to help you.

There are 5 most important default reports, that you can never miss on to dig down further and identify the data points like drop off of the users in their journey towards the final conversion, The performance of an individual product or product categories into the conversion or the source/Medium or Campaign that are driving the most conversions and much more.

Check the analysis that you can make out of those reports and how to use that as follows:

Shopping Behaviour Report:

Access: Conversions → Ecommerce → Shopping Behavior

First of all, having overall knowledge of the key touchpoints of the user journey is very important in order to identify the gaps and potential areas of improvement to increase the conversion rate. This identification starts with analyzing the journey of users and key drop-off points while carrying the users from the acquisition stage to the final conversion stage.

The Shopping Behaviour report plays a key role in identifying the key touchpoint from where the user abandons the website before converting. The report covers journeys starting from total user visits to product detail page visits and at last transacting on the website.

In the above screenshot, you can see that the maximum number of drop-offs is happening between the product detail page and add to cart stage, out of which the mobile device is showing the highest drop-off.

Now, the questions still remain, apart from overall website drop-offs, what else can this report answer. Well, don’t worry. The Shopping Behaviour report can also help us with the following questions.

Which device is performing worst on which step?
Which channel drives visitors to the first step, but not much further?
Do we have any browser issues at a certain step?

Once these pieces of information are available, you can create remarketing audiences to increase conversions which act as the first step for increasing conversions.

Product Performance Report

Access: Conversions -> Ecommerce -> Product Performance

While analyzing the touchpoints from where users tend to drop off from the website, it is very important to get the information of the product level granularity. In order to increase conversion rate, it is necessary to optimize the eCommerce store by measuring the performance of products viewed and purchased, product list views visited the most by the users, along with the optimization of the website experience, so that users get the most out of each journey done on the website.

In the above screenshot, as you can see that only a handful of the products are generating the maximum amount of revenue. It’s great news that we are aware of the top-performing products, it is also important to be aware of the low performing products and by checking the Cart-to-Detail Rate and Buy-to-Detail rate columns you can see which products are likely to sell well if promoted.

By analyzing these reports, you can devise further campaigns to show the desired products to users to increase the conversion rate.

Funnel Visualization Report

Access: Conversions -> Goals -> Funnel Visualization

Users go through a journey before buying something from your store and converting. This process, or path, can be visualised like a funnel. As there are many steps, there are many ways your funnel could have a leakage. This report gives you an on-point view of the conversion path and the data of each step.

This report will not tell you why they exited but it’ll give you a clear view of where they exited.
You can always use this information to know what exactly happened before they exited the page.

Check the content
Look for bugs
Check for technical problems

Or anything that might create friction in user experience while buying.
Don’t just start at the first step, go step by step - starting from the place in the funnel that has the lowest users on the page and proceeding forward.

Landing page, Source medium and campaigns report

Access: Behavior -> Site Content-> Landing Page

Access: Acquisition -> All Traffic -> Source / Medium

Access: Acquisition -> All Traffic -> Campaign

As you all know that all the eCommerce merchants are continuously running multiple campaigns simultaneously in order to increase brand name, product visibility to the multiple audiences and increase transaction volume as the key objectives. Although you can understand the overall website performance and product performances from the above reports, measuring the campaign performance also plays a key role in improving the conversion rate.

But, how to measure the performance of the campaigns run for each brand?

This can be done by analyzing three main reports i.e. landing page report, source/medium report and campaign report.

With the use of the Landing page report, You get an idea of the first-page user visited on your website/App and how many of them finally converted and the amount of revenue that they generated.
Use source/Medium report to get an idea of the source/medium which is driving the most traffic to the website and influencing the conversions and revenue.
With the help of a campaign report, you get to know the effectiveness of your campaign and the revenue and conversions that the specific campaign generates.

Site search reports

Access: Behavior -> Site search

Being an eCommerce business, your website/App will be having in-app/website search functionality. To track the performance of the search terms which are driving users to engage more on your website/app or driving more conversions, you can do all of it through site search reports available in GA.

With the help of this report, you get insights on how much time users spent, which pages are visited or directly exited after writing a specific search team. You also get an idea of the number of visits with or without site search or how many times the search refinement was done after the user search and so on.

The Bottomline is…

Google Analytics is a wealth of data available for you to increase conversions. While the reports we’ve mentioned here aren’t the only reports you should see but they should be at your top priority. There are a plethora of metrics and reports available in GA and you should track metrics that align with your business goals.
These reports can help you make data-driven decisions that will certainly provide much-needed momentum to your conversions.
To know more on how to optimise GA for your eCommerce store, get in touch with one of our representatives today!

Guide to Create CTA buttons to improve E-commerce value proposition

Posted on April 1, 2021 - 6:27 am by tatvic

5 Best Ways to Create Effective CTA Buttons and Effective Value Proposition for e-Commerce

What is a Value Proposition

A value proposition demonstrates the benefits your prospects get when they buy from you. It depicts the value you provide to your customers and gives customers a reason they might want to purchase from you. Value proposition answers an important question that each eCommerce customer wants to know from you: Why should I buy from you and not from your competitors?

Your value proposition is the main thing you can show on your website - if you get it right, it will bring an enormous boost. Value proposition details how you solve major pain-points your customer have and how you can improve their experience by adding more value to it.

The best value proposition should be clear on what is it and for whom? And How it helps.

Ways to Create Effective CTA Button for eCommerce

1. CTA [Call to Action] needs to be clear and Understood in five seconds or less.
Get straight to the purpose. Cut every unnecessary word, for example - Compare Buy Now with Add to Cart. One is far more urgent than the opposite. And how about replacing Try it for Free with a Free Trial. One is far punchier than the other and stands out effectively.

Do not ask for too much commitment and stick with one primary CTA, or if you are using more than one, then the secondary CTA can be a Ghost Button, but make sure the primary CTA remains highlighted.

2. The CTA should answer the customer’s principal need.

When it comes to your website, no element can be deemed as a part of a vacuum. It’s surrounded by other elements of varying sizes, importance, and messages, and jointly, all of them play a crucial role in determining the success and failure of that one element.

The same goes for the call-to-action buttons. What goes in the remaining part of the page, plays a big role in the effectiveness of the CTA buttons.

Also, focus on what customers get, instead of what they lose. For example, Use Get it Now, Free Shipping, Instead of Buy for $100 or Cancel Anytime, and Add To Cart - Save 25% instead of Buy Now.

Identify the highest motive for people that come to your website. The homepage CTA should address this motive and help move visitors towards their goal. People want to understand the maximum amount possible of something before they buy it, whether it’s about dimensions, features, or anything.

Users must be shown what they will get back in exchange for taking this action. To grab the user’s attention, play with powerful words and craft a highly compelling copy for your CTA button.

Using something attractive and unique instead of simple CTA like Submit Now, Shop Now, Buy Now, etc., will surely catch the user’s eyes and compel them to take action, but some powerful and unique words should be incorporated to get the users attention and get them more inclined towards the CTA button.

3. Stand Out with Contrasting Colors.

Colors do matter when you ought to achieve a huge conversion rate. Colors often won’t have much effect to assist the balance and the dimensions of your buttons. A highly effective CTA ensures an excellent use of contrasting colors that stands out from the rest of the background.

For larger buttons, choose a color that’s less prominent within your design but still stands out against the background. There’s no exact answer about which color works best. So, you’ve got to use and test different colors to examine which color can offer you better conversions.

For a smaller button, you may want to decide on a brighter, contrasting color to make the button pop. In either case, confirm the color you utilize, set the button apart without clashing with the site’s overall design.

So you’ve got to use and test different colors to ascertain which is performing best and provides better conversions, also it all depends on the design of your website. While doing so, make sure that it still resonates with the general look and feel of your brand.

4. Make sure the CTA is in the right size, place, and shape.

The CTA must be large enough to grab the visitor’s attention quickly and also make it easy for the user to click on it since the shape of the CTA also plays an important role in getting the conversions.

Depending on the design of your site, any shape might work well for you. Most e-commerce websites use rectangular buttons to make them more visible.

Also, keep in mind that CTA which is just too big will overpower everything around it. A small CTA will get lost in the shuffle of the other content on a page. So make sure that your CTA is large enough to face out without overwhelming the overall design.

As we know, using the mobile app to purchase or shop online is increasing, you must make the mobile experience as seamless as possible, which means that the CTAs need to be put where the thumb can easily reach.

Make sure that the user can click mobile CTA, with the thumb, by making them the entire width or nearly the entire width of the screen, simply it must be easily reachable to the user’s thumb where they can click with less effort.

Spacing matters a great deal, and to balance the amount of negative space you have around the CTA buttons so that it doesn’t look cluttered because negative space helps your CTA stand out from the rest [Especially on Mobile].

On pages for where you have a single desired action, remove any other unnecessary distractions. Note that this is less important on some buttons like Add to Cart buttons, with others, like those to Know More, works better with more space.

5. Prioritize and Create a Sense of Scarcity

The call to action (CTA) on the page must prioritize if they are more than one. Done in a few ways, but yes, the most common is by using color and size.

You can use colors to highlight the CTA button on a page or make them less important or less prominent. Also, resize the button to stand out by making it large.

Use simple, direct language and large bold fonts. Make sure the language calls for specific action.

width= Including visual cues in a CTA button is also an excellent idea, this helps to increase the conversion rates. The icon of a shopping cart - like the Add to Cart button, or an arrow on a download button, are both examples of visual cues. Make sure that the icon adds to the user experience by clarifying what the button is for, and doesn’t add any kind of confusion. Don’t be afraid to use less-commonly used icons, as long as their meaning remains clear.

Invoking a sense of scarcity can help. Show a deadline for a scheme or lack of stock, and by doing this, you can make the users more inclined to shop for your products. Create a sense of scarcity and urgency, by words like Hurry, Quickly, Limited Edition, etc.

When users know that this stock is limited, they rush to grab the product instantly, hence ultimately leading to an increase in conversions. For example: Showing sales end dates next to the CTA’s to remind customers that they have limited time to act. You can use this type of urgency-building element.

You want your buttons to offer them the impression that they have to act directly. You want to encourage them to make their decision immediately, Urgency is needed but doesn’t give false information or mislead the user, it will become a reason to pause.

In The End

The final tip for high conversion CTAs is the A/B Test and Optimize approach. The abovementioned tips will help you in boosting your conversions, there is no fixed approach for success in an eCommerce store, by conducting regular optimization tests, you will make incremental positive changes that will have a long-term impact than simply getting into blind.

The solution is to keep experimenting with variants of CTAs, copies, shapes, colors, etc., and find out what works better for you.

Do not forget to Incorporate a Value Proposition to help you get an instant boost in conversions by stating the value proposition in the call-to-action button.

How To Increase Your Advertising Effectiveness?

Posted on March 25, 2021 - 10:56 am by tatvic

Focus on Reaching The Right Person, At The Right Time, With The Right Message

Poorly targeted ads that reach the audience at the wrong time; compromise the overall effectiveness of any ad diluting the message that reaches the audience. An advertising nightmare!

Tatvic devised a unique approach and named it the Large Scale Experiment Method (LSEM) - to resolve the issue of poorly targeted ads, and LSEM increases the effectiveness of online advertising.

Brand identity is deeply affected by the kind of advertising and creatives used. It can build the affinity of people towards your brand or can appear repulsive.

Like most advertisers, if you want to increase your chance of creative success by reaching the right people with the right message at the right time, then this brief write-up about LSEM can help. It covers key points of LSEM and its role in developing creatives.

Large Scale Experiment Method (LSEM) is a method to reach the right person, at the right moment, with the right message by practicing the principles of persuasion and value proposition.

What is a Large Scale Experiment Method?

Large scale experiment method is a way to know your audience better. It provides actionable insights about the people with whom you want to communicate, what interests your audience, and be considerate of what and when they want to hear from you.

Harness the power of Data

You can leverage data to understand your audience. And figure out how you can sell to them or help them while entertaining them with your ad, that conveys your message most effectively. LSEM uses principles of persuasion and value proposition. So you harness the power of your data with accuracy and effectiveness.

Why LSEM?

If you are not getting the result you aim for, and your messages are not reaching the right audience, then pause and consider why?

There are some key questions to look upon to increase the effectiveness of your ads. Like - Have you paid much thought and consideration in the drafting of your creatives?

Are you able to get useful and actionable insights from your data?

Ineffective advertising messages that fail to gain attention might indicate a lack of consideration of the immense benefits of using data to generate useful insights. We will tell you how to do that by LSEM and in a way that helps you reach your audience when they like it.

Large scale experiment method helps draft non-intrusive messages by understanding the user’s interest. People might even appreciate you for being considerate about their interests, needs, and wants as it signals that you pay attention to them, listen and care enough to give the product or service of their choice.

Large Scale Experimentation Method (LSEM)

LSEM (Large Scale Experiment Method) is a method to Reach the Right person, at the right moment, with the right message by using the Principles of Persuasion and Value Proposition.

Method To Deliver More Relevant Creative

1. Identify The Moments

Identify the moments in which you want to engage the target audience.

To know your audience and understand what they are doing? And why are they doing it? You will need to understand the platform or channel where they are spending time? And why do they prefer that specific channel?

Once you have an understanding, what they are doing in these areas, you will be able to engage better.

Essentially it is all about reaching the right audience, at the right time and the right place. So you can more effectively know who your audience is by using targeted advertising.

Whether it is Google ads or social media ads, all of them offer advanced targeting options to reach your target audience. Also, you can target the ads based on the location, demographics, and interests of your audience.

Make sure that your ads are displayed only to people who would be interested in your brand. You will be saving your spending on ads and need not spend a fortune to reach your target audience.

2. Leverage Signals

Leverage signals to customize creative messaging.

Reach further with upper-funnel audiences to Increase your campaigns. If your main goal is brand awareness and you need help to expand your Display campaigns’ reach, consider using upper funnel segments, such as Detailed Demographics, Life Events, Affinity, and Custom Affinity.

Detailed Demographics: Detailed Demographics helps you reach users based on facts like - Homeownership, Marital status, Educational status, and Parenting stages. And it is not limited to only the age and gender of the user.

Life Events: Knowing and understanding the main transition points in your audience’s life, allows you to reach them at their crucial decision-making time when they are most likely to make important purchase decisions. Like when they are purchasing a home, starting a new job, and retiring.

Affinity and Custom Affinity: Affinity segments give specific insights into your user, their interests, and what they are passionate about?

These insights predict and depict your audience’s shopping behavior, lifestyle, and consumption habits.

Affinity audiences are mostly people who are frequent at shopping, dining out, visiting salons, or attending live events.

Similar audiences: Find new customers by building a model of your retargeted audience behaviors and characteristics and then applying them to potential new customers.

Brief of Signal types:

Audience signals: Types of people you’re trying to target.

Media signals: Info about the content your user is looking at, where your ads might appear.

Environmental signals: External factors that might influence the user’s mindset and behavior when your ads are displayed

3. Creative For Full Customer Journey

Develop creatives for the full consumer journey.

The customer journey is a detailed outline of every step a lead takes to become a paying customer, or we can say customer journeys elaborate the path of sequential steps taken and interactions made by the customers with a company, product, and/or service.

4. Test & Learn Marketing

Implement a more fluid process & adopt a test and learn approach. Test and learn marketing, is a data-driven approach to continuously improve your campaigns, website, and brand performance through a series of ongoing tests.

Chances are you already familiar with conversion rate optimization (CRO). CRO is a perfect example of test & learn marketing.

Turn insight into action from all of your tests that helps you to make informed predictions and decisions for future campaigns, for example :

Selection of the most effective ad format for your campaign.
Deeper insights into how you can segment your target audience to determine relevant messaging techniques.
See the performance of your creatives against your target audience lists.

This test and learn approach works on any marketing strategy. And automation tools make the task of managing tests and compiling data easier, that too without draining your funds or resources.

Benefits of LSEM

1. Engages an audience

To engage an audience, you would want to make smart use of your audience data. Audience data can include both first- and third-party data sources. Audience signals personalize campaign messaging. And make it more relevant based on the insights about the audience and their interests.

2. Saves time

No more mapping keywords, bid, and add text to each product on your website. Plus, By changing the category in the feeds, it is easy to target a particular audience.

To save you time, and keep your ads relevant, Dynamic Search Ad headlines, landing pages are generated using content from your website.

3. Shows relevant ads

Google Ads dynamically generates, shows relevant ads, with a clear and concise headline for the most relevant page on your site. This ad is generated based on the relevancy of the customer search with your product or service.

4. Control your ads

You can show ads based on your entire website, or specific categories, or landing pages. Prevent your ads from showing for products that are temporarily not performing well.

You can easily customize your ads by knowing which elements in the ads are not performing better. For example, CTA text, copies, and Images to improve your ad performance.

Conclusion

No brainer, your creatives reflect and affect your brand identity and your brand communicates through it. The ads should be less intrusive, well-timed, and reach the right person conveying thoughtful and well-crafted messages.

You get the crux here - Reach the right person, in the right moment and through the right message, and BOOM, your ad creative wins admiration from your audience, and you will rightfully gain their attention. Be remembered for better.

Send your queries our way at he***@****ic.comto know more about LSEM based Digital Advertising services and how can it amplify your digital advertising. You can also check a few examples of our Principle-based, Rich media, Interactive, Data-driven creative advertising here.

How to Track App Unintals in IOS & Android using Firebase? - Part 1/2

Posted on June 15, 2017 - 9:49 am by tatvic

Blog App Uninstall Feaure Image

App uninstalls are the easiest thing to do on our smartphones, right? Enough blogs have mentioned the reasons behind these app uninstalls and a multitude of companies have done surveys too, reporting to reach at 10 worst reasons causing app uninstalls. For e.g. Google App Marketing Survey says an average app user has 36 apps installed on his/her smartphone. But they use only one-fourth of these apps daily. The remaining one-fourth that are never used are the ones likely to be uninstalled. Placing few other links in the references section.

A Kantar/ITR study showed that an average of 26% of app installs are uninstalled in the first hour. That uninstall rate rises to 38% in the first day, 64% in the first month, and about 89% over 12 months. Those figures represent the average across all app types.

App Marketer’s KPI woes:

If you are a fellow app marketer or a product manager like me, you can feel the pain of a leaky bucket of users uninstalling your app mercilessly. We toil day and night to ensure a sprint app development process, user-friendly app features, and carve the simplest user journeys. We then market it just enough to make our app stay on top of Google Play Store rankings. All seems well and we are right on track to achieve that glorious milestone of 1 million app downloads. But our app analytics tools like Google Analytics, Firebase, Apsalar have a completely different story for us.

The report suggests that app uninstalls are on the rise this month, and your (well, mine too) KPIs for “number of active installs” indicates a churn rate of more than 40%.

What’s the Next Plan of Action?

Armed with the intelligence of the analytics tools, we decode the behavioral attributes of the users/ user segments who have uninstalled your app. The reasons behind most of the uninstall become our next app optimization and development tasks. But we almost forgot about the churned users. How do we win them back?

The first instinct is to take resort in our run-of-the-mill marketing activities.

We run mailer campaigns with offers/marketing gimmicks if they had logged in with their email ids
We send promotional SMS to user inbox if we had managed to capture their phone numbers.

And say, a majority of these users never logged in or gave their personal details to us. Now, as a consequence, we can never retarget them with personalization and the only thing left for us to do is just wait for these users to reinstall our app.

What’s the solution? Do we have a way to stop the users from churning in the first place?

Predictive Action to Stop Users from Churning?

Machine Learning Prediction Model is the answer. At Tatvic, our data science team has come up with a prediction model which feeds in scores of users, devices, and behavior features from the analytics tools and Tatvic’s internal Uninstall Library and predicts the probability of users who will uninstall within the next ‘n ‘ days. We call it PredictN Model.

Here is how it functions - in brief - we input the cohort of users who were acquired between any 15-day period along with their attributes. With previously trained data of identified churned users, PredictN Model will point out unique users from the said cohort who have more than 75% probability of uninstalling within 7 days.

Retarget these users at once via Paid channels and Push notifications
Understand the pattern of these unique users as to why they might be leaving your app from a geographic, demographic, device-specific, and/or behavioral perspective.

How does the model work?

Check this space to know more about model attributes and results in our next blog chapter.

A takeaway for reading till the end - heartfelt thanks, Feel free to reach out to us in case of any query or leave a comment in the section below on how you take predictive actions for uninstalls in your business.

Sources:

Predictive analysis on Web Analytics tool data

Posted on July 3, 2013 - 7:17 am by tatvic

In our previous webinar, we discussed on basic things to perform predictive analysis. We also discussed on an eCommerce problem and how it can be solved using predictive analysis. In this post, I will explain R script that I used to perform predictive analysis during webinar. I will also highlight how web analytics support can enhance data accuracy and streamline the analysis process for better business insights.

Before I explain about R script, let me recall the eCommerce problem that we discussed during webinar so can get better idea about the data and R script. For eCommerce retailers product return is headache and higher return rates impact the bottom line of their business. So if return rate is reduced by a small amount then it would impact on the total revenue. In order to reduce return rate, we need to identify transactions where probability of product return is higher, if we can able to identify those transactions then we can perform some actions before delivering products and reduce the return rate.

In webinar, we discussed that we can solve this problem using predictive analytics and use Google Analytics data. To perform predictive analysis we need to go through modeling process and following are the major steps of it.

Load input data
Introducing model variables
Create model
Check model performance
Apply model on test data

I have included these steps in R script. So, let me explain R script that we used in webinar. R script is shown below.

# Step-1 : Read train dataset
train 
# remove TransactionID from train dataset
train 
# Step-3 : Create model 
model 
# Step-4 : Calculate accuracy of model
predicted 
#Step-5 : Applying model on test data
#Load test dataset 
test 
#Predict for test data
test_predict 
#creating label for test dataset
label 
# set label equal to 1 where probabilty of return > 0.6
label[test_predict>0.6] 
# attach label to test dataset
test$label 
# Identify transactionID where label is 1.
high_prob_transactionIds 
high_prob_transactionIds

As you can see that first step is load input data set. In our case input data are train data and train data are loaded using read.csv() function. Train data contain the transaction based data and it contains TransactionID. TransactionID is not needed to use in the model, so it should be removed from the train data.

We also discussed about the variables during the webinar. Train data include pre-purchase, in-purchase and some general attributes. We can retrieve these data from the Google Analytics.

Next, model is created using glm() function and three arguments are given to it which are formula, family and data. In formula, we specify response variable and predictor variables separated by ~ sign. Second argument we set family equal to binomial and last we set data equal to train. Once model is created, its performance is checked where accuracy of the model is calculated. it is shown in the script.

Finally, model is applied on the test dataset and predict the probability of the product return for each transaction in test dataset. In the script, you can see that I have performed several steps to identify the transactionIDs from test data having higher probability of product return. Let me explain them, first test data are loaded. Second, predict() function is used which will generate the probabilities of product return and store in test_predict. Third, new variable label is created which contain 0 for all transactions initially and then using test_predict variable, 0 is replaced with the 1 where probability of return is greater than 0.6 or 60%. Now this label is attached to the test data. Finally all the transactionIDs are retrieved where label is 1 which means that probability of product return is greater than 60% in these transactionIDs.

So this is the script which I used during the webinar and performed the predictive analysis. I have created dummy datasets which you can use to perform these steps yourself. You can download data and R script from here

Here I want to share you one thing, this is not optimized model. This is a practice model for understanding how to leverage web analytics support services for your business needs. You can improve the model by taking other variables from Google Analytics or performing some optimization tasks, so you can get better results. However if you want to look at some other predictive models on web analytics tool data click here

Google analytics data extraction in R

Posted on December 3, 2012 - 6:49 am by tatvic

Unlike other posts on this blog, this particular post is more focused on coding using R so an audience with a developer mindset would like it more than pure business analysts.

My goal is to describe an alternative method to use to extract the data from Google Analytics via API into R. I have been using R for quite some time but I think the GA library for R has been broken and while they did make an update, it’s sort of not being used right now.

Considering this, I thought to write it down by myself on moving on as more data-related operations are now being done using R.

Moreover, the Rgoogleanalytics package that is available is built for Linux only and my Windows friends may just like me to have something for them as well.

Ok so let’s get started, it’s going to be very quick and easy.

There are some prerequisites for GA Data extraction in R:

At least one domain must be registered with your Google analytics account
R installed with the following the Googleng packages
- Rcurl (download by clicking here)
- rjson (download by clicking here)

Steps to be followed for Google Analytics data extraction in R :

Set the Google Analytics query parameters for preparing the request URI:

To extract the Google Analytics data, first, you need to define the query parameters like dimensions, metrics, start date, end date, sort, and filters as per your requirement.

# Defining Google analytics search query parameters

# Set the dimensions and metrics
ga_dimensions <- 'ga:visitorType,ga:operatingSystem,ga:country'
ga_matrics <- 'ga:visits,ga:bounces,ga:avgTimeOnSite'

# Set the starting and ending date
startdate <- '2012-01-01'
enddate <- '2012-11-30'

# Set the segment, sort and filters
segment <- 'dynamic::ga:operatingSystem==Android'
sort <- 'ga:visits'
filters <- 'ga:visits>2'

Get the access token from Oauth 2.0 Playground

We will obtain the access token from Oauth 2.0 Playground. Following are the steps for generating the access token.

Go to Oauth 2.0 Playground
Select Analytics API and click on the Authorize APIs button with providing your related account credentials
Generate the access token by clicking on the Exchange authorization code for tokens and set it to the access token variable in the provided R script

Retrieve and select the Profile

From the below, you can retrieve the number of profiles registered under your Google Analytics account. With this, you can have the related GA profile id. Before retrieving profiles ensure that the access token is present.

We can retrieve the profile by requesting to Management API with accesstoken as a parameter, it will return the JSON response. Here are the steps to convert the response to the list and store it in to the data frame object profiles.

# For requesting the GA profiles and store the JSON response in to GA.profiles.Json variable
GA.profiles.Json <- getURL(paste("https://www.googleapis.com/analytics/v3/management/accounts/~all/webproperties/~all/profiles?access_token=",access_token, sep="", collapse = ","))

# To convert resonse variable GA.profiles.Json to list
GA.profiles.List <- fromJSON(GA.profiles.Json, method='C')

# Storing the profile id and name to profile.id and profile.name variable
GA.profiles.param <- t(sapply(GA.profiles.List$items,
                              '[', 1:max(sapply(GA.profiles.List$items, length)))) 
profiles.id <- as.character(GA.profiles.param[, 1])
profiles.name <- as.character(GA.profiles.param[, 7])

# Storing the profile.id and profile.name to profiles data.frame
profiles <- data.frame(id=profiles.id,name=profiles.name)

We have stored the profiles information in profiles data frame with profile id and profile name. We can print the retrieved list by following code

profiles

OUTPUT::

         id       name
1 ga:123456    abc.com
2 ga:234567    xyz.com

At a time we can retrieve the Google analytics data from only one GA profile. so we need to define the profile id for which we want to retrieve the GA data. You can select the related profile id from the above output and store it in to profileid variable to be later used in the code.

# Set your google analytics profile id
profileid <- 'ga:123456'

Retrieving GA data

Requesting the Google Analytics data to Google analytics data feed API with an access token and all of the query parameters defined as dimensions, metrics, start date, end date, sort, and filters.

# Request URI for querying the Google analytics Data
GA.Data <- getURL(paste('https://www.googleapis.com/analytics/v3/data/ga?',
                        'ids=',profileid,
			'&dimensions=',ga_dimensions,
                        '&metrics=',ga_matrics,
			'&start-date=',startdate,
                        '&end-date=',enddate,
                        '&segment=',segment,
			'&sort=',sort,
			'&filters=',filters,
                        '&max-results=',10000,
                        '&start-index=',start_index*10000+1,
                        '&access_token=',accesstoken, sep='', collapse=''))

This request returns a response body with the JSON structure. Therefore to interpret these response values we need to convert it to a list object.

# For converting the Json data to list object GA.list
GA.list <- fromJSON(GA.Data, method='C')

Now its easy to get the response parameters from this list object. So, the total number of the data rows will be obtained by the following command

# For getting the total number of the data rows
totalrow <-  GA.list$totalResults

Storing GA data in a Data frame

Storing the Google Analytics response data in R dataframe object which is more appropriate for data visualization and data modeling in R

# Splitting the ga_matrics to vectors
metrics_vec <- unlist(strsplit(ga_matrics,split=','))

# Splitting the ga_dimensions to vectors
dimension_vec <-unlist(strsplit(ga_dimensions,split=','))

# To splitting the columns name from string object(dimension_vec)
ColnamesDimension <- gsub('ga:','',dimension_vec)

# To splitting the columns name from string object(metrics_vec)
ColnamesMetric <- gsub('ga:','',metrics_vec)

# Combining dimension and metric column names to col_names
col_names <- c(ColnamesDimension,ColnamesMetric)
colnames(finalres) <- col_names

# To convert the object GArows to dataframe type
GA.DF <- as.data.frame(finalres)

Finally the retrieved data is stored in GA.DF dataframe. You can chek it’s top data by the following command

head(GA.DF)

OUTPUT::
        visitorType operatingSystem   country visits bounces      avgTimeOnSite
1       New Visitor         Android Australia      3       1              106.0
2       New Visitor         Android   Belgium      3       1 155.33333333333334
3       New Visitor         Android    Poland      3       0               60.0
4       New Visitor         Android    Serbia      3       2 40.666666666666664
5       New Visitor         Android     Spain      3       1               43.0
6 Returning Visitor         Android (not set)      3       3                0.0

You will need this full R script for trying this yourself, You can download this script by clicking here. Currently, I am working on the development of R package, which will help R users to do the same task with less effort. If anyone among you is interested provide your email id in the comment, and we’ll get in touch.

Would you like to understand the value of predictive analysis when applied to web analytics data to help improve your understanding relationship between different variables? We think you may like to watch our Webinar - How to perform predictive analysis on your web analytics tool data. Watch the Replay now!

Predict User’s Return Visit within a day part-2

Posted on October 22, 2012 - 7:18 am by tatvic

Welcome to the second part of the series on predicting user’s revisit to the website. In my earlier blog Logistic Regression with R, I discussed what is logistic regression. In the first part of the series, we applied logistic regression to available data set. The problem statement there was whether a user will return in the next 24 hours or not. The model is built and till now it was showing us 88% accuracy in predicting user’s revisit.

In this post, I’d try to showcase ways to improve this accuracy and take it to the next level. This is more about technical optimization so if you are a business reader you may want to skip and check how can you use this for your benefit. But, if you are techwiz or Data modeling guy like me, let’s get rolling.

As I have discussed in blog Improving Bounce Rate prediction Model for Google Analytics Data, the first step of the model improvement is variable selection and the second step is outlier detection (If you want to know more details of steps, refer mentioned blog). Let’s apply these steps one by one.

Variable selection

I have used stepwise backward selection method for variable selection. R code for the stepwise backward selection method is as below.

>Model_1 <- glm(revisit ~ DaySinceLastVisit + visitCount +f.medium +f.landingPagePath +f.exitPagepath+pageDepth, data=data ,family = binomial("logit"))
>library(MASS)
>stepAIC(Model_1, direction="backward")

Output
Start:  AIC=2119.37
revisit ~ DaySinceLastVisit + visitCount + f.medium + f.landingPagePath +  f.exitPagepath + pageDepth

                     Df Deviance    AIC
- f.exitPagepath    152   1732.4 1966.4
- f.landingPagePath  87   1751.0 2115.0
                    1581.4 2119.4
- pageDepth           1   1583.4 2119.4
- f.medium           11   1656.5 2172.5
- visitCount          1   1740.1 2276.1
- DaySinceLastVisit   1   1826.4 2362.4

Step:  AIC=1966.42
revisit ~ DaySinceLastVisit + visitCount + f.medium + f.landingPagePath + pageDepth

                     Df Deviance    AIC
                    1732.4 1966.4
- pageDepth           1   1738.9 1970.9
- f.landingPagePath 101   1987.5 2019.5
- f.medium           12   1821.2 2031.2
- visitCount          1   1929.3 2161.3
- DaySinceLastVisit   1   1978.4 2210.4

Before we understand the output, let me explain how the variables are selected in stepwise backward selection? In the stepwise backward selection method, AIC is used as the selection criterion. General rule is lower the AIC, best the model(i.e. For a group of variables, if AIC decrease by removing any variable(s) from group,then remaining variables are used in the model. This process continues until AIC stops decreasing). From the output, we can see that AIC is decreased and variable exitPageapath is excluded from the model. Now, we will create new model(Model_2 ) which does not include exitPageapath. R code for new model is as below.

>Model_2<-glm(revisit ~ DaySinceLastVisit + visitCount +f.medium +f.landingPagePath +pageDepth, data=data,family = binomial("logit"))

After generating the new model ,let’s check the accuracy of the new model and it is as below.

>predicted_revisit<- round(predict(Model_2,in_d,type="response"))
>confusion_matrix<- ftable(revisit, predicted_revisit)
>accuracy<- sum(diag(confusion_matrix))/2555*100

Output
86.57534

From the output, we can see that accuracy of the new model is decreased. This does not seem good to us. Variable selection method did not help us in improving the model. Let’s try second step for model improvement which is outlier detection.

Outlier detection

As we know that data set contains some unreliable observations which make model’s quality poor. We always need to detect outlier and remove them. For numerical variables, outliers can be removed by observing the histogram of frequency distribution of the values of each variable (Process is described in blog Improving Bounce Rate Prediction Model for Google Analytics Data). In our data set, there are three numerical variables named visitCount, daySinceLastVisit and pageDepth. I have generated new data set after removing outliers. Let’s create new model based on new data set and check the accuracy of the new model. R code for new model is as below.

>Model_3 <- glm(revisit ~ DaySinceLastVisit + visitCount +f.medium +f.landingPagePath +f.exitPagepath+pageDepth, data=data_outlier_removed ,family = binomial("logit"))

Now, we will check the accuracy of the new model and it is as below.

>predicted_revisit <- round(predict(Model_3,in_d,type="response"))
>confusion_matrix <- ftable(revisit, predicted_revisit)
>accuracy <- sum(diag(confusion_matrix))/2292*100

Output
98.42932

From the result, we can see that model has more accuracy than previous models (Model_1 and Model_2) and it is good for us. So, removing the outliers from the data set, the model got more improvement and prediction accuracy. For now, we can conclude that through this model (Model_3), we can predict more accurately whether a user will return to website in next 24 hours. If you want to do exercise, Click here for R code and sample data set. In the next blog, we will discuss about logistic regression with Google Prediction API, check the accuracy of the Google Prediction API for our data set and try to predict for a user that will user return to website in next 24 hours?

Would you like to understand the value of predictive analysis when applied on web analytics data to help improve your understanding relationship between different variables? We think you may like to watch our Webinar - How to perform predictive analysis on your web analytics tool data. Watch the Replay now!

prediction api

Posted on October 3, 2012 - 5:40 pm by tatvic

In this post, I am going to explain how can we build model for transactional product revenue prediction application with Google Prediction API as we already discussed same stuff (Product revenue prediction with R)on R. With the help of Prediction API, we can build prediction model without any programming. Here we just have to focus on our dataset to make more accurate predictions. Google provides this service with Google Cloud Storage and Google Prediction API.

We can store our dataset at Google cloud storage and fire queries from Google prediction API for generating prediction on sets of your dataset which is easy. Google Prediction API is developed with set of optimized Machine Learning Algorithms for model based prediction. Here is description on the base of the Prediction API v1.5

How to train data with Google Prediction API:

First, we require minimum one number of project at our Google API console, enabled Google Prediction API and Google Cloud Storage service with that project. Our datasets have to meet certain format like

There must no header for all column
Predicted variable must be at first column
There must not NA values in dataset which confuse Google Prediction server

Then we can upload our dataset to Google Cloud Storage Engine by creating bucket. After uploading dataset to Google Cloud Storage Engine, we have to query from Google Prediction API by Google API explorer. There are numbers of methods to Prediction API service to deal with the data like

prediction.trainedmodels.insert - Can insert your dataset for training with this function.
prediction.trainedmodels.get - Can get the training status of pre-inserted data model
prediction.trainedmodels.analyze - Can provide the Analysis on the trained model
prediction.trainedmodels.predict - Can make prediction

As in previous blog, we have already developed model in R for same dataset. For doing same with Google Prediction API, We can start training of dataset with prediction.trainedmodels.insert with model unique id, data storage location, and type of model as parameters.

Will give Response like:

It means request accepted by Google Prediction server and have started training. We can check the training status of model with the help of prediction.trainedmodels.get

Here is response of Get function of Prediction API:

Where it describes

Toatal numbers of instance = 4061
Type of model = Regression
Mean squared Error(MSE) = 1606123.17

Here, this model summary will not provide intercept and variable coefficient like R. To make prediction with above model we can query by Predict method (prediction.trainedmodels.predict)

Test 1:

xcartadd = 0
xcartuniqadd = 0
xcartaddtotalrs = 0
xcartremove = 0
xcardtremovetotal = 0
xcardtremovetotalrs = 0
xproductviews = 47
xuniqprodview = 38
xprodviewinrs = 5828

Actual (Productrevenue) = 110.06

 Output
(Productrevenue) = 155.15717487938028

Test 2:

xcartadd = 0
xcartuniqadd = 0
xcartaddtotalrs = 0
xcartremove = 0
xcardtremovetotal = 0
xcardtremovetotalrs = 0
xproductviews = 484
xuniqprodview = 392
xprodviewinrs = 445026

Actual (Productrevenue) = 803.81

 Output
(Productrevenue) = 934

with R (after removing outliers + subsets of independent variable), we predicted 115.8346013 for test 1 and 955.153476 for test 2 on same dataset. Therefore, getting prediction with R and Google Prediction API are nearly same. When we are doing with prediction API we can improve prediction with improving dataset quality but another side in R, we can improve prediction accuracy by improving dataset quality as well as prediction model as we already discussed in Product revenue prediction with R - part 2

Hello There!

Author Archives: tatvic

About tatvic