Posterous theme by Cory Watilo
Vimal Tripathi

Vimal Tripathi

My thoughts on Business, Technology and the Business of Technology

More about me @ about.me/vimalt

TwitterBlogger

Cohort Analysis For Marketers

I touched upon the concept of cohort analysis in my previous post. I would like to exapnd a bit more on cohort analysis in this post. There is enough content available about the analysis itself, but most of it is focused towards UX designers or developers. In this post I want to share my thoughts on the extend utility of cohort analysis, especially for marketing. Let’s start with the basics first.

What is Cohort Analysis?

It is an analysis of how a group of individuals (or objects), that share a common characteristic, behave over time. In a business context, it is used to analyse the behaviour of customers through their life cycle. There are many grouping criteria that can be used to perform cohort analysis. The most common grouping criterion that online businesses take into consideration is the joining date of the users, grouping them by date, week or month.

Why should one use it?

Consider the diagram below. The first graph shows the number of new users added per week by a particular product. The second graph shows the product usage of customers grouped by their month of joining. The second graph shows that after two weeks of product usage, 50% of users never come back to the product. By week four, 50% of the remaining customers abandon the product. While the first graph shows growth in the business, the second graph shows lack in engagement.

Growth_vs_engagement

If the business is growing, the overall usage of a particular feature may seem to be on the rise, however, cohort analysis will filter out growth from customer engagement and tell you whether users are actually engaging with the product the way you’d want them to. This is further explained nicely by Fred Wilson and by 52 weeks of UX

Cohort Analysis for Product Marketing:

So you have done a great job at creating awareness and you’ve got customers who have signed up for your 60-day free trial. How do you now make sure that at the end of the 60-day trial period these customers are willing to pay for your product, in short how do you increase your monetisation rate? This might be a good time to look at the user data and do some cohort analysis.

As a marketer you want to give users the information they need to recognize the value that your product provides, especially when users are on their on-boarding journey during their first few days with the product. Analyse what are the first few features that a new customer uses, let’s say first five features. If one sees a consistency of this usage of features across cohorts, it becomes clear which features customers access when they start using your product. Marketers can use this information to tailor the on-boarding emails they send to new customers, thereby reducing the barriers to uptake of the product. This data can be combined with another cohort analysis to get even deeper insights. Just like we chronologically ranked feature usage, it will also be insightful to rank these features according to the amount of usage. If you see a similarity in the pattern of this data, you would know which top 5 features are most heavily used by your customers in the first 60 days. You can now combine these two cohort analyses and build a type of perceptual map e.g. two axes based on urgency and importance. The figure below illustrates this with an example of an email service.

Imp_vs_urgent

This data can again be used to build your email communication, self help tools, FAQ’s, enabling support etc. You can use a similar analysis to find out what are the last few activities that the customers do before they abandon/cancel the product usage. If this is consistent within cohorts, this might give you a good idea about the point in their life cycle at which the customers could do with a little help.

Cohort analysis is also a great way to ascertain the life time value of a customer. By looking at the historic data of individual cohorts, average lifetime value of users can be calculated. This data can be a goldmine for finance, as it can help them get better with forecasting and budgeting.

Like any analysis, cohort analysis cannot be used in isolation. One has to maintain a good balance of numerical data and feedback from users to understand their individual pain points. As someone once said:

“Happy families are all alike; every unhappy family is unhappy in its own way.”

10 things I learnt from Evernote and Phil Libin

I love applications that are simple, easy to use and that really solve problems. For these very reasons I love Evernote. Evernote has got simplicity, seamless multi-platform access, and the best part, it's free. In this post I am going to share with you some of the notes that I took while listening to a talk delivered by Phil Libin, CEO - Evernote (terrific guy)

1. Evernote doesn't want you to hear about Evernote from Evernote.
Now that's something. This philosophy is based on the belief that if you hear about Evernote through an ad, or a promo email, the chances of you ignoring it are far higher than if you heard about it from a friend who says "Dude! There is this cool app I found called Evernote". They want their existing users to share the goodness of the product with their friends. They believe that referred users have a higher retention rate than the ones that are not referred. Their closed Beta for the product had a tremendous response rate as the new users were recruited only through friends.

2. If the product is non scarcity based, let people use it for free.
According to Phil, "The easiest way to get 1 million people paying is to get 1 billion people using". If you are an Evernote user, you may have already observed this: there is no real barrier to be able to enjoy Evernote free, lifelong. Neither does one receive any kind of marketing push to monetise to paid version; their marketing is busy doing something else that I will share with you shortly. They want to get as many people using the product for free; of course they want to achieve a monetisation conversion rate too. Surprisingly, they don't want to see that conversion ratio increasing beyond a point, as that would mean that they are not pushing enough free users into the funnel.

3. Analysi Cohortem is the magic spell.
For subscription based business models, cohort analysis is the most important instrument at hand to gain insights into customer data and usage patterns. Cohort analysis is the analysis of monitoring behaviour of a group of users that share a common characteristic; in this case the cohort is formed of users that joined Evernote in a particular month.  For example, cohort analysis for Evernote shows: 
  • consistency in the amount of software usage it takes for users to stay with the product. Once that barrier is crossed, users continue to use the product forever. This time-barrier is consistent across cohorts 
  • consistency in the time after which users start to monetise 
  • a pattern in reduction in variable cost/user/month 
  • a pattern in increase of avg. rev./user/month 
Snapshot view of businesses like X revenue, Y unit sales, Z ARPU etc is very relevant in conventional business models. They are still relevant in SaaS business models, but due to a fundamental difference in revenue recognition in subscription based models, cohort analysis gives much deeper insights into the business, forecasts and most importantly, user behaviour.

4. We have a life time to make money from you.
Hasn't it happened to all of us that we find this amazing restaurant with an amazing dessert, however, we realise that over time it doesn't taste as nice as it tasted the first few times. Compare that with a newspaper subscription that you received in inheritance from your granddad. Unless the newspaper starts deteriorating in its quality your perceived value of the newspaper doesn't change much over time. Now imagine you are using Evernote regularly for 6 months, the perceived value of Evernote for you would change tremendously from the day you started using it, to a month of usage to 6 months worth of usage. If you are a heavy user, after 6 months you will practically have 6 months of memory in Evernote.
Perceived_value

What we have just seen is three kinds of product/services whose perceived value changes differently with time. In Evernote's example the perceived value of the product increases the more/longer you use it. This is precisely the reason why Phil and the gang don't send me glossy marketing emails. They are happy with me using Evernote for free. So long as I am using it, they have my whole life to 'monetise' me.

5. There is no better way of spending money than spending it on the product. 
Phil says, if you make an awesome product, the sheer goodness of the product takes care of word of mouth. These guys spend most of their money building the product and packing it with a punch of absolute awesomeness, rather than spending huge amounts of money marketing it. By the way, Evernote does have a marketing team, continue reading further to know what exactly they do. 

6. You can't get it cheaper than free. 
Freemium coupled with a great product can do wonders. This is particularly important for startups, as the only way the Microsofts and the Googles of the world can compete with you is by making a better product than yours and not by undercutting you; it is difficult to be cheaper than free. Plus it is fairer to lose to a better product than to a cheaper one.

7. How big is your target market?...7 billion people?
Who doesn't want a little help with their memory. I think team Evernote have done a great job at making an application of which target market is pretty much anyone who uses a computer/smartphone/tablet, working/home maker, Mac/Windows/Android/Blackberry/Palm/Chrome/IE/Firefox user . Oh! by the way, you can even use Evernote on your conventional paper notepad using Livescribe. 

8. It can be used in more ways than you thought.
This is where it becomes clearer what do the marketing folks at Evernote do if they don't make quirky ads or glossy emails. Evernote marketing helps its users by showing how Evernote is being used by other users. Team Evernote itself was unaware of the various applications of their product and the way people were using it. So their marketing teams set out to discover the different ways in which people all over the world are using Evernote: students using it to maintain lecture notes, shoppers using it for creating shopping lists, travellers creating photo journals, chefs using it to create their recipe journal. The possibilities are endless and Evernote marketing acts as a facilitator helping users help other users.

9. Geek Meritocracy is here.
It's the time of the Geeks. App Store + Cloud Computing + Open Source + Social Media + Freemium = Geek Meritocracy
Geek_meritocracy

10. Libin’s Law = Moore’s Law x Murphy’s Law
Just to end on a funny note. Here is Libin's Law, which is a blend of Murphy's Law and Moore's Law. According to Phil Libin, the number of things that will go wrong at a startup will go double every year.

While Libin's Law seems humorous, it is not distant from the truth. The company almost came down to closure due to cash crunch. Until one night a Swedish user sent a note to Phil saying that he absolutely loved the product and would be happy to help if the company needed money. $500,000 later they got back on their feet and got back to doing what they do the best - making an amazing product. And here I am writing about them in Evernote, using the notes that I had made in Evernote!

 

ZMOT: Zero Moment of Truth

Recently, on a friend's recommendation I read a book called Zero Moment of Truth. In this book Google (the publisher) talks about the changing buyer behavior and the importance of online marketing. In many ways this book looked like a subliminal catalogue of Google's offerings telling marketers to spend more money on search ads...and some more money on search ads. That said, the book does reiterate some of the already existing awareness around online marketing. But it is worth having them all at one place. So here are my key takeaways from this book.

Conventionally, marketers have believed in a three-step consumer journey- Stimulus, Shelf and Experience.

  • Stimulus occurs when the consumers become aware of a product via TV, radio, print ads. 
  • Shelf is the time when consumers venture out into the stores and are exposed to the actual product along with many other competing products to choose from (P&G called this the First Moment of Truth or FMOT or eff-Mot). 
  • Once bought, consumers experience the product, and depending on whether they are satisfied or dissatisfied, they share their experience with friends, colleagues, neighbours etc. Today, things have changed. Between the Stimulus and the Shelf, sits a new moment called the ZMOT or the Zero Moment of Truth (coined by Google) or Zee-Mot.

Zmot_-_zero_moment_of_truth

In this day and age where the internet follows us pretty much everywhere (PC/Laptop/Tablet/Phone), consumers are always looking for information about products and services the moment they hear about them through a ‘stimulus’. Consumers are looking for information in every step of buying process:

  • At the consideration phase when they are looking for answers whether they need a product?
  • When they are comparing products.
  • Looking for reviews the moment they consider buying a product. Experiences that others have had with the product.
  • Choosing from where to buy the chosen a product? Looking for Offers, discounts and coupons.

Consumers these days are so aware that they walk into a showroom and it is likely they will say “I want to drive the Lexus 250h with the touring package, the 2.4-liter four-cylinder engine, dual-zone climate control and tan leather with seat warmers. I’ve seen the specs and read the reviews online. And I know that the MSRP is $37,125, but your real invoice price is $33,686.”(Excerpt from ZMOT)

People search for reviews on everything from paper clips to pizzas to penthouses. And the people looking for reviews are everyone from expecting moms to base jumpers. People are sharing their experiences with not only friends and neighbours, but with complete strangers on review websites, online forums etc.  Word of mouth (WOM) is stronger than ever and it happens one-to-many millions. Not to anybody’s surprise, this is what WOM looks like online.

  • Consumers talking directly to millions over social networks, YouTube videos (They even have product un-boxing videos), comparison websites, review websites, online retailers and so on.
  • Ratings and comments that appear with business names on applications like Bing Maps, Google Maps, Kayak, augmented reality mobile apps.
  • Interest groups and community websites.
  • Ratings on search ads like Google ads

That being said, most companies are skeptical about providing rating feature on their product pages. Understandably, they are afraid of any negative commentary on their own product pages. But the truth is, reviews are happening, whether one likes it or not. So why not do it on your own website and know who the detractors are. In my experience, if one can provide that extra bit of service and take corrective measure to please the detractors, they eventually become your biggest advocates. Moreover, some negative commentary only makes rest of the positive reviews look authentic.

While a review is a consumer making a conversation about your product, what do marketers need to do to seize this opportunity? Even today, many businesses think that building a slick website is their answer to creating an ‘online presence’. Long gone are the days when creating a website was your ticket to creating a presence online. Marketers need to use the web as THE most important medium to converse with their customers.  Setting up a website without any engagement with consumers is like setting up a 1-800 number with nobody to answer the phone. People are reaching your website from various sources, marketers need to join the conversation, on the website or their Facebook page or elsewhere on the internet, and engage with customers. To do a quick check on the word out there, try searching these on Google:

  • [Your brand name]
  • [Your brand name] review
  • Best [your category]
  • Start typing your product/brand name in Google and see what auto-fill options you see

If you don’t like what you see, your prospects won’t either. Brands need to be there when their prospects are looking for answers related to their product. When consumers are looking for answers, give them the answers before you push the product. Because the truth is that the prospect has not made a decision yet and all they are looking for is help to make that decision. Help them in making the decision and more often than not you will end up being the considered product.

Enough said about the importance of ZMOT. Now some words on how to win at ZMOT. Answer the questions that consumers are asking. Analyse what questions people are asking before they make a decision. Answer those questions first and then push the product. If I searched on a search engine "What are the ingredients in a curry sauce", I would expect to be shown the ingredients of the sauce than knowing whether I can get "10% off curry sauce" on your website; I have not made the decision yet, so help me in making the decision. Optimise your website with relevant content. To be a thought leader within your category, you need to build content around educating consumers. Be their go-to resource and help them in making the buying decision, before selling your product to them. Create how-to videos, FAQs, whitepapers etc. And again, optimise your website!!!Finally, be agile. Test and Learn.

And remember, while doing any of the above, fail faster. The faster you try, the faster you may fail, the faster you may learn and faster you may succeed.