Funnel Optimization

(Roberto Bolanos ) #1

Which articles essays you specifically recommend for funnel optimization?

The things I’m struggling:

  • should I do simultaneous experiments in different parts of the funnel?
  • why yes or why not
  • how many can I
  • should I start at the top or at the bottom?

(Andrew Chen) #2

for just the landing page, here’s a great book:

I agree that generally there’s a missing part to talk about funnel optimization more generally. But I haven’t found great resources besides just talking to folks who’ve been working on this area for a long time.

(Roberto Bolanos ) #3


I never thought you would personally answer my question. That was REALLY awesome.

Thank you really much! I just bought the book. If you have other suggestions for funnel optimization s please let me know

(David Janczyn) #4

Hey Roberto,

My new job is Direct Response funnel heavy:

  1. Depends, because I’ve seen tests increase clickthroughs on Step 1 but then hurt conversions downfunnel. Because of this, I usually run 1 test per main funnel, and am shoot for more holistic impact, vs. just click throughs on a step. However you can run a multitude of ‘post-purchase’ tests on the backend (upsells, referrals). We have loads of paid traffic, so my cadence is 1-2 weeks max per main test, which enables fairly high velocity and iteration
  2. I’d start at the very bottom, checkout page. Those are the highest intent visitors.

(Roberto Bolanos ) #5

Great Advice David.

I will start with the bottom, and test one piece at the time with the holistic approach!

Please keep in touch. I’m a serious student of Direct Response. I think you will love “Making Advertising Pay” full of cases of Master Lester Wundermann.

If you have any experiments, books, essays you recommend please let me know here or at

(Laura Levy) #6

I really recommend this article on conversion funnels:

In addition to that, here are some general tips for funnel optimization:

  • Measure time between each stage - is it taking an abnormal amount of time for users to get from one step to another?
  • Look for bottlenecks where users get stuck and make improvements around those areas of the product
  • No matter how many tests you end up running, make sure to look at the effects each change has on the entire funnel, and not just whether they converted to the next step
  • One more super useful hack: watch user session recordings to understand the user experience that led to the dropoff

Good luck!