You showed me your uniques, your page views, your downloads. You showed me your transactions, and your growth rate. You regaled me with examples of Unicorns (for the un-initiated: start-ups with more than $1 billion valuation) who inspire you. You told me how you’ll be worth a billion dollars if you stay on the Path.
You explained to me, as if you were speaking to a child, about the amount of money some of the best venture capitalists (VCs) had put in a stealth company just like yours in the Valley a couple of months ago.
It was all very impressive.
Your month-on-month growth was greater than 30%, app downloads were ramping up wonderfully. You had, of course, decided to be mobile-only (I disagreed but kept silent).
Oh, revenues. Well, revenues would come automatically, wouldn’t they? All we had to do was get ahead of everyone else and become the biggest game in town. Then a mystical virtuous cycle would kick in—and we would all be rolling in more money than we could imagine. In any case, who cares about revenues? If we have 25 million customers, we’ll be worth X billion dollars based on the going rate. Valuation metrics the world over were $30 per monthly active user weren’t they?
I listened. We can always wait for revenues. Sometimes they do come in the end if you do everything else right.
And then I asked you about your cohorts—I wanted to see them.
I asked you for simple evidence that you are actually solving a problem for consumers, that consumers actually need you and want to deal with you. Proof that they are willing to stretch out and meet you halfway.
You showed me.
And I *shivered*.
There was no evidence.
It was like a perfect crime.
Your monthly repeat rate was lower than your growth rate. Your referral rate could be counted on the fingers of one hand. Your conversion rate was below a percent. I was aghast.
My face must have shown it.
But you went on. You told me how you had become more efficient at marketing; your customer acquisition cost had dropped by 50% over the last two months as you had managed to learn how to hack Google Adwords and Facebook. How you had found a nice affiliate partner who was getting you app downloads at less than Rs 13 a pop.
I told you that there was no hack in these metrics, it was just plain rigour on analytics tools that a school-going child would have applied. But somehow the irony was lost on you.
So, I had to pull you back to reality—that hacking Adword buying, or bribing customers with deals to download apps or to buy was not what the business was about.
That’s how we started chatting about the heart of the business, about your reason for existence and your Truth.
And I started telling you about the lessons I had learnt from the time I first started making internet investments in 1999.
Seems a long time ago now, but Truths don’t change, just how we interpret them does.
There are only three Truths to your business:
#1 How many of your customers repeat
#2 How often do they repeat
#3 The rest of your business is just a support system for #1 and #2
Depending on your specific business, you can substitute ”repeat” with any of the following: share, convert, like, fill a form, refer, sign-up, time spent, pages viewed, ratings, review.
Tweak it to fit your circumstances, whether you are fulfilling a constant need, a one-time purchase or are a content business designed to buy a consumer’s attention.
VCs have a label for it: “finding product-market fit” (PMF). It’s a nice euphemism for “do people really need your s**t”.
Why do they focus on this so much?
PMF is a marker for a stage in their investment. It’s an important marker since all investments prior to this stage are based on faith in the team, and a general belief in the proposition. But when they see signs of a start-up achieving PMF, that’s the night they sleep well. That’s when they are happy to double-down and put in more money into your business.
Till that point it’s like spraying small bullets in the dark. Once you find PMF, they want you to pull out the rocket-launcher and press fire.
That’s their moment of validation; yours was when you closed the funding round.
To make PMF a bit more real for the purpose of this piece, I am going to call it a “cohort”—you can apply any of the labels I’ve used above, or create ones that suit your business.
If your cohorts are weak, you have a problem.
Q: How did the start-up fail? A: One cohort at a time
You can hide it under massive user growth, snuff it with millions of downloads, cover it under some great PR around your fundraise.
But when it comes down to the atomic level—when it is just between that single customer and you—there is no place to hide.
Demographically and psychographically (ouch, what a mouthful), India probably has the most heterogeneous audience in the world. Language, age and just the sheer new-ness for customers using their smartphone for the first time are things that you are dealing with. New instruments, fear of dealing with online, paranoia of being cheated, fear of making a mistake and being charged for it, are all barriers to a smooth transaction. Add to it the perennial struggle with internet access and download speed.
So where does one begin improving cohorts:
It starts with a realization that it is not just a repeat rate. It is repeat rate from this market, this gender, this device, this prior purchase, this offer, this source.
You need to analyze each of these situations separately. Go back to your school days. Start making sets of these cohorts, intersect them, visualize them as people—don’t let them be numbers on an Excel sheet.
Work backwards, and attempt to discover or at least imagine the intent behind their actions. That will give rise to meaningful insights.
If you’re a pure-play internet company, your business is simply the culmination of your consumers’ micro behaviours—each tiny one matters. (And if you are not pure-play, you need to factor in your employees’ micro behaviours as well.)
Times have changed. We are entering a phase where how well you run your business will count more than the size of your capitalization table (which shows the shareholders’ stake and how much they paid for it in each round of funding). So here are some tips to get your start-up to deliver on its promise.
The Data hides more than it reveals
This is a hard-learned lesson from my years in the media business. You should be vitally aware of the “self-fulfilling” error when you analyse your cohorts—some things end up working precisely because of how you place them in the flow. Sometimes the product or that headline or that offer is clicked more often because it was put on top of the page, and not necessarily because people were interested in it.
I call this in-built error the “Anti-Heisenberg Principle”: position can actually determine the momentum! More seriously, it causes huge errors in site design and planning as data analysts focus on consumption without seeing how placement in the mix is forcing that variable.
Your design and usage data is its own enemy. Consumers behave along the path you set them. Don’t blame them; it’s like putting trains on a track that goes off a cliff, and then wondering why the data is showing the trains are crashing.
Don’t blame the train, blame your track!
Data analysis is not all science, it’s quite a bit of art. The link between causation and correlation is flawed and it needs your gut and instinct.
A few weeks ago, I had a meeting with the founder of a fairly successful, growing, bootstrapped private label start-up. He had fantastic organic traffic—millions of visitors, great time-spent, but very low sales.
The repeat visitors and referral metrics were good but strangely conversion was low. Clearly, people came to the site with an intent to purchase, but something was killing it.
We discussed what do consumers typically see. He told me that for every category he offers over 250 options and keeps adding new ones every week. I asked him to visualize from their point of view. Are you confusing them with more options?
Four weeks later he sent an email saying he had cut his options by over 30% with absolutely no loss of sale but a massive saving in backend and inventory costs. He plans to cut down the options further and create unique paths for first-time and returning customers.
He now intends to personalize the mailers, and create follow-up mailers with products that either complete an ensemble or contrast it for the consumers and is getting his tech backend ready to create customized mailers depending on his customers’ purchase pattern.
Lesson: Get your analytics/intelligence team to put away their laptops, call in for some beer and sit and have a chat—not about what the data is showing, but to wonder what people could possibly be doing. Ask them to run a series of “what ifs” and create new train tracks. The results will surprise you.
And yes, tell them to step out of their cubicles and go buy a coffee for customers (and non-customers) in the real world.
Every nuance matters
There are hundreds of case-studies on how a small change to the call-to-action from green to red caused a massive jump in conversions. Scores of A/B studies (experiments with two variants) talk about the endless small tweaks that flipped consumers over from just browsing to buying.
They are all true. Isn’t it obvious? The only thing between you and your customer is that tiny 4x5-inch mobile screen, that 12-inch laptop screen—each bit of that real estate and how you treat it matters.
Colours, position, verbiage—every tiny cue matters. The success of Slack proves that even your attitude matters. You can become a boring intra-office application (several of them tried and died), or you can innovate on every little feature and cue and redefine the market like Slack is doing.
Just a couple of quick examples on how these nuances matter:
When I launched Moneycontrol.com, we realized that Indian consumers hated setting up portfolios on financial websites. Way back in 2001, we crafted a feature that would pre-populate a portfolio based on their search pattern (using cookies). That small tweak ensured that Moneycontrol has never been challenged as India’s premier financial site (despite minimal innovation for the last five years).
I recall a debate with Bookmyshow.com’s Ashish Hemrajani and his team in 2007 (Network18 invested in the company). We decided to remove the pain of signing-in to buy a ticket. We asked why should we need a sign-in? Our job was simple: get the customer to his ticket with the least roadblocks. It’s a very small tweak and most of you wouldn’t have noticed it yet, but you keep coming back to Bookmyshow for its smooth transaction experience.
Similarly, I see most travel sites don’t keep a record of my past booking history—that I usually book only direct flights (by the way, this hopping flight concept is not relevant to India; it originated in the US, where airlines fly hub-and-spoke), or that I always give priority to departure time versus hunting for the lowest fare. Yet these sites treat me like a brand-new customer and insist on showing me two-hop flights between Mumbai and Delhi from 2 am in the morning.
It’s amazing that in this day and age, several websites and apps struggle to treat me differently from a new customer. If 100-odd ad-tech companies can know my usage pattern, why is your website still in the dark? Find a fix!
Bought traffic means nothing after a while
If your customer acquisition cost (CAC) is not lower than your 6 or 12 month Lifetime Value (or LTV, the amount you earn from a customer over that period), you are likely to be toast.
You are probably spending all the money, all over again, to get the customer back.
If you do the math at your start-up, you will discover that you make money only from heavy users.
We spend all our money and effort on generating trial and not enough on the “follow through”—on the journey of converting trial users into heavy users.
Light users look very rosy on VC presentations since they show up in the monthly active users and daily active users metrics, but they are harmful, as they suck away resources without ever converting to loyal customers.
Split your audiences into the first timers, light users and heavy users. If you are able to shift just 5% of your light users to heavy, you can change the fortunes of your business. For instance, if 15% of your users are heavy users, that tiny 5% more can give a 33% jump in your revenues without adding to costs. That’s your leverage, right there.
Many a start-up has died at this altar of widespread adoption but thin repeat usage; consumers come for the deals and leave you holding the lemon.
There is a tipping point in the consumers’ behaviour when they convert from flirting users to heavy ones. Get obsessed with finding it.
If your monthly growth rate is constantly higher than your repeat rate, you are one cohort away from failure.
Your visitor is a flesh and blood person with a motivation
Don't get carried away by the testimonials from Google Play Store or the reviews from your Facebook page. Wake up and smell the coffee—this is feedback from just 1 or 2 % of people who used your service and not from the 98% who never transacted.
The silent majority votes with their feet; they just click away from your carefully designed website, your mouth-watering deals and don’t come back.
If you don’t ask them, you’ll never know.
You need to invest time and effort researching your non-users. Find a way to get into their head. We’ve been reacting to selling cues for thousands of years—it’s visceral. Observe saree salesmen at Delhi’s Karol Bagh market and you’ll see how they change their behaviour once they size up a new customer walking into the shop and ensure she leaves with more sarees than she ever intended to buy. That’s the way you win at your business.
Often your team will be obsessed with adding features just because they think it’s clever. They forget that there is some poor consumer at the other end of that maze struggling to get your app to do something simple and doesn’t really need these fancy little features you have added. Empathize with him!
Over-engineering is a crime. It’s just crazy to assume that consumers want more and more and more—don’t overwhelm them with choices and options.
Example: Not a single e-commerce player has created a simple usage and buying guide for the millions of first-time consumers. There is no content to help people choose their TV sets or smartphones. The strategy seems to be to just overwhelm them with 250 models and bribe them with a discount. In five years most haven’t figured that their site can be a “favourite” place to shop if it actually helps consumers make a decision and not just blind them with options.
How many e-tailers have tried to use the power of video that can crush the language barrier? Why are most e-commerce sites in English? Somehow they are willing to spend millions on discounts and nothing on creating a hook. Why shouldn’t YouTube be flooded with Flipkart’s or Snapdeal’s buying guides? Even the comparison shopping sites haven’t tried to carve an identity for themselves.
Lesson: Don’t put your team in silos when it comes to consumer empathy. Have a regular “flow" meeting. Ask your entire team—tech, product, merchandising and sales support—to sit and figure how a consumer would react to each touch point.
This meeting is more important than the one with the affiliate marketer at the cafe next door.
We are not good at design
There, I said it. Let’s admit that we Indians have been introduced to good design only in the last few years. We were not born into it, not seeped into it. We haven’t been reading London Tube maps for the last 50 years and haven’t seen great road and highway signage since birth.
Try most of the packaging from Indian consumer companies and you’ll see what I mean. Some of them are sealed with such an intensity that you wonder if they ever wanted a consumer to unpack them. Only recently have we started seeing some innovation in packaging and design.
We have a cultural heritage of art and aesthetic design, but not functional and simple design. Though I’ve worked with some of the best, the Great Indian Designer is a rarity. So if you have one, keep him—else find a way for your team to collaborate with the best in the world. Those are the best dollars you may spend. Like it or not, most Indian designers who will make the leap for you are still in high school.
Some founders also forget the essential difference between a user interface designer and a user experience designer. They are not the same thing.
They shouldn’t even be sitting together. One lives in the studio and the other should be out in the field with your customers.
It doesn’t end with your website or app. It merely starts there. The end of every transaction is the beginning of a new, deeper relationship with the customer.
It’s a fresh chance to reinforce your engagement, to take trust and the bond to another level. Every moment of delight you create matters. And this gives rise to a completely new set of cohorts for you to analyze.
Cohort analysis is the fundamental underpinning of your business. It is the lie-detector test for whether your model works. It indicates, whether you set the right expectations, and whether you meet them. Are you creating productive, useful customer relationships?
Good, solid cohorts are the indicator that your model is differentiated and sustainable.
In other words, they are why you come to work.
Remember, you are one cohort away from failure
Oh, oh, now you tell me that she downloaded the app because you were giving her friend a Rs 1,000 coupon as a referral freebie. Hmmm. Now that’s quite another conversation. We need to discuss how you are spending your money. It doesn’t grow on trees, and there’s been, you know, this market crash in tech stocks. So, you know, it’s important that we discuss how you are spending your dollars. But that’s a conversation for another day….