jtbd optimize

When Funnels only Explain a Minority of Purchase Scenarios

You’ve been modelling your buyers into a funnel.

  1. XXXXX visitors land on the site
  2. XXX visitors going to the product page
  3. XX visitors sign up for a newsletter
  4. XX visitors go to trial
  5. X trials convert to a purchase

The problem though, is that this funnel only explains, like, 15% of your conversions.

15%! What explains the other 85%?

You’ve got a model, but the model ain’t the whole story.

Your Model Leaves Questions Unanswered

You dig into your analytics, but you can’t find the answers.

  • What’s the rest of the story look like?
  • Who are those other people that drop-off the path?
  • Should I even care about those other people?
  • How come I have people who are happy with the newsletter, but never signing up for a trial?
  • Are people being turned off by something? (confidence taking a hit)
  • Is my product a “no” or a “not yet”

Multiple funnels describing multiple pathways? Behavorial analytics and heat maps? You know you could go in that direction.

But at the stage you’re at, your analytics alone can’t be modelled to give you the whole story, not with your constraints. You don’t have enough traffic to run experiments. You can’t wait months to get behavorial analytics to show some clues.

But there’s a reason you’ve stayed in analytics-land, up till now. That’s because you’ve seen mixed results going to the other side: the qualitative side didn’t have enough hard truths.

Getting Qualitative, Only With Actual Reality

There are a bunch of qualitative tools we won’t be using to get answers about your buyers.

We won’t be using survey results. No focus groups either. No list of feature requests nor lists of customer interactions.

You want to stay away from those because they all have one difficult bias to overcome: they all have asked your customers to use the creative side of their brains. For that reason alone, those aren’t grounded in reality enough.

What you want to get to are retrospective accounts of what happened leading up to the purchase. You want to hear (five to ten) purchase stories.

Purchase Stories Get You To Reality

Purchase stories are retroactive accounts of the events that lead up to a buyer making a purchase. From a purchase story, you’ll get the two key events that created momentum toward a “yes, this, now”, beating the gravity well of the “I’ll just” statements, your product’s real competitors. You’ll find the forces that were pushing and pulling in the mind of the buyer.

Those aspects aren’t made up. They’ll give a dose of anectodal (not statistical) reality into your understanding of your buyers.

Are 5 to 10 interviews enough? Yes. I’ll explain why in a minute.

Here’s how you get purchase stories:

  1. Find some recent buyers who hesitated to buy your product (a screener questionnaire will be needed, the only kind of survey that will be helpful in the process) along with a few other important factors for qualifying a buyer to be interviewed.
  2. Offer an incentive for those who are selected.
  3. Interview them using some retroactive questions (this is the important bit), which bring back the buyer into the events that led up to the purchase. You want questions that encourage recall and dig for motivation.
  4. Record the audio for later, with the participants permission, so you don’t have to take as many notes, and so you can have others in your team listen to the purchase stories.

You’ll Start To Uncover The Whole Picture

After 5 to 10 interviews, you’ll start seeing the “job” for which they really “hired” your product:

Stuff like that.

More importantly, it’ll help you draw a better picture of your funnel.

Explaining The Holes In The Funnel

Better stated, purchase stories will help you see how each one of the steps in your funnel were answering different jobs-to-be-done, which might explain why your funnel didn’t really flow.

  1. Your website was “hired” to get answers for free.
    Nothing to do with your product.
  2. Your newsletter was “hired” for the job of not missing any of your insightul perspectives.
    Nothing to do with your product.
  3. Your product page and the trial were “hired” to put a proposal together for a boss to greenlight.
    Nothing to do with your product.
  4. Finally, your product was “hired” for a very specific job despite all the features you were touting which caused the buyer to hesitate (“Will I use all these features?”).

The Re-Model That Might Ensue

What we’re looking for is job fit, and layered jobs.

  1. Job fit: Each element you publish needs to answer a job-to-be-done.
  2. Layered jobs: Each higher-value element needs to answer a job-to-be-done a lower-value element creates.

This way, you might not have a single funnel, but multiple purchase pathways. And that’s fine.

  • Maybe create an information product teaching the process that your SaaS product helps automate
    Same overall direction, two jobs-to-be-done, one feeds on the other.
  • Maybe create a best hits collection or a guide for a topic your newsletter talks about.
    Same overall direction, two jobs-to-be-done, one feeds on the other.
  • Maybe pare down the features of your product to solve the main job-to-be-done really well and create a smaller, paid tier. Make sure your higher-tier product answers jobs-to-be-done that you only get once you use the lower tier.
    Same overall direction, two jobs-to-be-done, one feeds on the other.

So no need to count on home-runs. Have several at-bats.

No need to try to feed each interaction into that one funnel. No need to minimize drop-offs at all costs.

Just make sure you know the job-to-be-done each thing you have is hired to do, make them solidly answer those jobs, and boom. Value-ladder.

With those insights, you’ll be itching to go back to modelling, back to analytics. This time, you’ll come back with a new dose of reality.

Stay Sharp!

—

@pascallaliberte

Some articles that answer the "job" of knowing if this approach is legit:

Get new articles like this one.

To learn to sharpen your sales pages.