Make User Projections That Mean Something

(Update: May 2017) This post originally appeared in 2014. We’ve updated it with what we’ve learned since then. 

Startups often join StartupYard right at the beginning of their journey. For B2C companies, having a few hundred or a few thousand MAU (Monthly Active Users), can provide a wealth of insights about what in the product works, what doesn’t, and what the users want from it.

But how does a startup with essentially zero confirmed users make decent User Projections based on little or no evidence?

Context

One of the key mistakes that B2C companies make is to see trends in their user growth that don’t actually hold up. Every B2C startup wants hockey-stick growth, but the less data you actually have, the less precise your projections based on that data will be.

If my user numbers doubled last week from 1 to 2, I can claim that my userbase is doubling on a weekly basis. Then I should reasonably predict that by week 26, I will have nearly 68 Million users!

Depending on your sense of what’s reasonable. Or I could as easily predict that I will have just 26 users. Regardless, I can be very sure that the real number will probably fall somewhere in the middle.

Startups need to be both ambitious, and realistic about their growth projections. This is essential to creating a plan that will actually achieve those results. In this post, we’re going to go through the process of establishing reasonable and falsifiable assumptions, that will help you achieve your user growth goals.

Working Backwards

Noah Kagan of Mint.com, discusses this process in his informative blog. He points out that the only workable approach to gaining users is to work backwards from a clearly defined goal, breaking down the channels and methods of user acquisition, their costs and their timescales, into one simple to follow spreadsheet. And while his own projections of how Mint.com would reach 100,000 users in 6 months seem wildly optimistic (including a 25% conversion rate from sponsored adds on Digg.com, with a CTR of 10%), this type of planning actually brought Mint a million users in that same period.

 

 

The Assumption Spreadsheet

Before launching, it’s important to relate your marketing goals to your assumptions. If your goal is, say, 30,000 users within the first six months after launch, you’ll have to justify that growth with something more than hope. You’ll have to show how you assume it can be done. Catalogue and quantify your methods of getting this growth, in ways that can be tested quickly and definitively, including the costs of acquisition per user.

Your assumptions for marketing costs might look something like this:

Source	Traffic	CTR	% Conversion	CPC	Users
Organic Search	100000	10%	25%	$1	2500
Native Ads	500000	10%	20%	$2	10000
Google Ads	1000000	2%	20%	$0.50	4000
Media	500000	5%	25%	$2	6250
Direct Marketing	1000000	1%	25%	$1	2500

Cost Total:  $39,500       User Total: 25,250

Projections are Not a Plan

Your assumption spreadsheet is more than just a plan for user acquisition, or a marketing plan.

This is just a part of a much larger set of assumptions. Other factors include your churn rate, your pricing, your internal growth costs, etc. But everything should relate back to basic assumptions that can be challenged and adjusted before launch.

Ask yourself: what will happen in 30 days, when I find that the CTR for one source is half what I expected? Alternatively, what if my CPC for one source is much lower than I predicted?

Going back to the table and adjusting the assumptions as data comes in, helps you to see where your time is best spent going forward. If CTR is low, maybe the message doesn’t match the medium, and you need to rethink your call to action. If conversion is super high somewhere, maybe you need to increase the spend there at the expense of other channels.

Be Reactive

Death to an early stage startup is a failure to react to data. I can’t count the number of times I’ve looked at early marketing data for a startup, and pointed out the above issues, only to be met with blank stares. You must incorporate changes in your data as often as possible, and iterate as often and as quickly as you can, in order to press advantages, and eliminate disadvantages.

“Wait and see” is death to a startup. You must react to every change in the data. Data must guide your planning.

Be Inquisitive

Sometimes you need to seek outside advice. Divorcing your assumptions from objective reality can be hard for anyone. That’s why mentors and advisors exist, so use them.

For example, my spreadsheet above has a few obvious problems. First, I’ve obviously fallen in love with the idea of “native advertising,” or the kinds of ads that fit in with the content of a given site. This makes sense, as I am personally a fan of this type of advertising. But by scrutinizing the spreadsheet, I can see that native advertising is predicted to cost me $40,000 out of a total marketing budget of $56,000, or 71% of my budget. Despite that, only 58% of my users will come from native ads. They will cost me a lot, and will not be as valuable as the users I gain from Google Ads, or even direct email marketing, according to my own assumptions.

Having your plan challenged can reveal your biases and hangups. Often, just sitting down and gaming out the plan for yourself can show you that you’re locked into the wrong thinking.

Have an Answer

It’s not necessarily wrong that I would spend 70% of my marketing budget on native ads, when they will only bring me 58% of my users. After all, users are not points of data in the end: they are people. Different kinds of users will behave in different ways. Maybe the CLV (Customer Lifetime Value), of a user acquired from a particular channel is higher than one from another. I can’t know that. The best I can do is assume, and test assumptions. 

These are the kinds of details that investors will pounce upon when they are shown your user projections. You need to have an answer for why you would pursue that avenue of user acquisition over another. The conversation has to be about your assumptions, not about what you don’t know. 

Perhaps these users are of a higher value (they buy more expensive products), or perhaps they are likely to stay with your product for longer. Perhaps the market for your product in Google ads won’t give you that same CPC if you spend twice as much on them. The size of the market may not justify a bigger focus on Google ads over native ads.

Don’t Fall in Love with Unknowns

Donald Rumsfeld, U.S. Secretary of Defense under George W. Bush, famously justified poor planning for the American occupation of Iraq in 2003 by saying: “There are known knowns, known unknowns, and unknown unknowns.” As many critics later pointed out, the Bush administration leaned heavily on the idea of “unknown unknowns,” to justify a lack of realistic long-term plans that matched their aims in the conflict. Anything that was inconvenient to think about or defend to the public became an “unknown unknown.”

The result was years of escalating conflict that have continued to plague the region to this day.

You can surely see the parallels between this kind of expensive, poorly planned and overly-optimistic strategy can be a cautionary tale in business as well. Refusal to undertake planning and spell out -and then recognize and react to- expected difficulties can lead you to do things that cannot be undone. You can undo a product change, but you can’t un-spend wasted advertising budget. You can’t un-launch a product that has failed to launch.

 

How Big is Your Mountain?

“How big is your mountain?” That’s the metaphor our StartupYard’s MD Cedric Maloux uses to describe this process of reconciling ambition with the need for some realism, even in the growth phase of a startup.

Ambitions to reach 1 billion users are great, if your product is the kind of thing 1 billion people can get some value from at the same time. Not many products are like that, which is why most don’t have the chance to grow that big.

But perhaps your mountain isn’t made of a massive user community. Perhaps it’s made of a smaller, quality user base, that pays a reasonable but significant amount to use your product, and gets a disproportionately large value out of it. In either case, your first six months or so should represent the first “summit” of your mountain.

Within that time, with the help of investors who understand and support your journey, you should reach a goal that is both ambitious and achievable, and you should gain valuable understanding of real users, and what it takes to make them happy.

 

Justify the Impossible : Control the Conversation

While the assumption spreadsheet functions as a roadmap for growth, it also works as a roadmap of your thinking for investors and partners.

Blind ambition, particularly when you’re asking for money, is not an attractive quality. But qualified ambition, in which you set hard goals for yourself, but also show how achieving those goals can be possible, can be very enticing to investors.

To use the mountain analogy: you can ask investors to help get you to the first summit. A goal they can believe in. That moves the conversation about the ultimate goal to a more appropriate time; a time in which you have more data and more momentum.

Imagine two founders who have essentially identical products. Both will meet with investors. One does not prepare these predictions because, as he says, “you just can’t know what will happen.” The other makes predictions, each representing his ambitious plans to achieve benchmarks in user acquisition. One of these two has something real to talk about with the investors. The other doesn’t.

How do you think those two conversations differ? I can tell you from experience: the one who comes with predictions gets enormous value form investor feedback about those predictions. The one without them gets, at best, idle speculation. If you aren’t willing to do the mental heavy lifting of making predictions, investors aren’t going to do it for you.

To define the conversation and lead it, you have to make statements that you can back up. Instead of having investors question your ideas, have them question your predictions. They’ll be speaking in your terms, rather than theirs.