A funnel is the basic building block of a startup. Every product feature is a funnel. Every marketing campaign is a funnel. Every upsell is a funnel.

You should analyze your funnels relentlessly. Case in point: Branchout’s tremendous growth is due to a relentless focus on marginal improvements of their funnel. Every day for 6 months, the company set small daily improvement targets in the conversion funnel. These small gains compounded to create exponential growth. And then? 25M users. Boom.

Below are three different funnels, A, B and C. Which one would you rather see in your business?

Funnel Visit % Convert to User User % Convert to Paid Paid Users
A 1,000,000 10% 100,000 0.50% 500
B 1,000,000 5% 50,000 1% 500
C 1,000,000 1% 10,000 5% 500

A is the best funnel. Why? It has the best performance at the top of the funnel (the  step from visit to user).

The top of the funnel is the most critical part of the funnel because:

  1. Fits the 3 main product priorities. Distribution, Engagement, Monetization. Those are your priorities in that order. Your funnel should reflect this focus.
  2. Proves product resonance. A 10% conversion to user is a strong indicator that you have identified a problem for your customers and marketed the solution in a way that resonates with your customers. A 1% conversion to user indicates the opposite: your message is unclear or your problem doesn’t resonate.
  3. Offers a much larger data set. Funnel A has twice the data points on usage as funnel B. More data is better data. Better data drives better decisions and faster time to reaching an optimal product state. Plus, you have twice the number of email addresses for user retention and re-engagement should customers churn.
  4. Prevents early optimization. Global maximums are better than local maximums. You want to make certain that you’re optimizing the funnel across all users. If you start with a small population of users and optimize the tail of the funnel for them, you may be leaving a much larger, more valuable group of users on the table. You won’t know until you have a larger data set.

So at the beginning, focus on distribution – the top of your funnel. Then engagement – the middle of your funnel. Then monetization – the end of the funnel. You’ll build more successful products.


7 thoughts on “Maximizing your conversion funnels

  1. Sure, as long as your variable costs are close to zero. Costs seem to be a vague concept which someday everyone (everyone!) gets the pleasure of dealing with

    • Most of the companies I’m seeing have a very low variable customer acquisition cost. They use social media (Facebook, Twitter, Pinterest) or mobile app stores (iOS, Android) for distribution rather than paid search. SEO is often a component too. But the principle applies even to online education companies like University of Phoenix which spends $4k to acquire a student. They maximize the top of their funnel provided the ROI is positive.

  2. Startups HAVE TO HAVE TO HAVE TO chip away at this every day, carefully adjusting/measuring/repeat each of these sliders.

    A slight edit though, in my experience –
    Create a ‘Sign-ups’ column between Visit and User.

    This will provide a granular focus to –
    Sign-ups – Product resonance, product packaging, signup friction
    Users – Engagement, stickiness, features

    This will also ensure that a sign-up is not considered a user rightaway, a very important discipline to maintain in the early years. You must define a user as a ‘real’ user eg.uses product atleast x times in y days

    Infact, even your excellent post on 3 priorities should include one more just before engagement – Sign-up (for the lack of a better term). You’d be surprised at the drop-off at sign-up!

    • Thanks, Rishit. You’re right – the more granular the funnel, the better you can measure and tune it. The distinction you make between a user and a sign-up is an important one for sure.

  3. I’ve also seen advice to optimize your funnel “in reverse” – ensure you are decent at converting to paid, then ensure you are good at converting signups to users, then ensure you are good at generating signups.

    I believe this reverse optimization approach is probably more capital efficient – if you get good at getting people into the front of your funnel before you are good at converting them to revenue, then there will be a period when you are spending time/money on those “front of funnel” activities without receiving any true traction/revenue benefits. This if fine if you have plenty of capital to work with, but not if you need to generate at least some revenue quickly because you are at least partly bootstrapping.

    • I like that advice when the funnel is more mature. The risk in doing the optimization in reverse is a local optimization risk – the funnel is optimized at the beginning for a small group of power users instead of a broader audience. In so doing, you might be maximizing revenue at the cost of distribution.

      But once a product is well established in the market, then I agree.

Comments are closed.