A step-by-step guide to using Outcome Driven Innovation (ODI)

For our new product Envato Elements, we convinced our boss to let us use Outcome Driven Innovation (ODI). I’m glad we did! For starters, we got vital insights into how customers measure the success of their job-to-be-done. Plus we then had the focus & confidence to make bold, product-shaping decisions.

I’ve heard that many ODI projects take 12 to 24 weeks to complete. But we didn’t have the luxury of that sort of time. Speed was of the essence here.

We successfully implemented Outcome Driven Innovation for Envato Elements in less than 6 weeks of part-time effort.

It took us about 2 weeks to prepare and run the interviews, and to consolidate that down to 94 customer outcomes for our job-to-be-done. After a short detour on another project, we then took another month to design the survey, get 500+ responses, analyze the data, and write up our findings.

All of this effort was part-time while working on other things. And using what we learned, I’m sure we could do it faster, better and cheaper next time…

A step by step guide to implementing ODI quickly

Here’s a guide to quickly and effectively using Outcome Driven Innovation, based on how we implemented it on our project.

1 Understand the Outcome Driven Innovation method. Read the book What Customers Want and the various Strategyn whitepapers. Map out in detail how to implement it. No need to tell your boss as yet (that’s best done later!) Brief key colleagues and enlist a few supporters who want the results.

Read it. Several times. Closely!

2 Work through an end-to-end mock analysis. Generate some fictional sample data like what you’ll get from surveys. Set up scripts (we used R) to process the export files from your survey software and to crunch the data. Tweak it until you’re happy that it outputs the right results, tables & charts.

Example output from our mock analysis

3 Define your job-to-be-done. Make it simple, clear and understandable. For example, ours was “Source digital assets for use on a creative project”. Map out the key process steps involved in that job. For us, that included item search, decision, purchase, item download, and account management.

Example process steps in our job-to-be-done

4Prepare templates to structure your outcome collection interviews and to make collecting outcomes efficient. Interview yourself and see how many outcomes you get. (We got 20–40 per interview). Refine the templates and build your skills further by doing mock interviews with colleagues.

A template we created to capture outcomes quickly in interviews

5Identify target interviewees you already know. Brainstorm people, scan your professional network, and ask colleagues. Call in favors to get their help ASAP with research on this new product.

6Bring a colleague along to the interviews. Run them in person if possible, but hangouts, phone, or skype are fine too. Get permission to record it. Use your templates to capture outcomes and key commentary as you go. Later on, get the audio recording transcribed for a more detailed record.

7Once you stop hearing many new outcomes in your interviews, that’s enough! Copy all the outcomes into a single spreadsheet. Consolidate them to remove duplicates with different wording. Get to a single list of around 100 customer outcomes for the chosen job-to-be-done.

8Prepare a summary of key insights captured in interviews. Have the full list of outcomes in a handy format. Create a couple of slides about the process. Prepare a draft survey design in an online tool (we used Survey Monkey). Decide on survey incentives and how to promote it.

9Meet with your boss, explain the basic ideas of ODI, and highlight some interesting insights so far. Emphasize that the analysis is ready to go, you just need the survey data. Oh, and $ for incentives. Go back and forth to streamline the survey so it’s as easy as possible to take… but still true to ODI.

An example section from our survey. Note how we used color, wording, layout, etc to make it friendlier & simpler.

10Launch the survey online. Send email to your offers list and invite people who meet the qualifying criteria to complete it. If you’re offering per-response incentives, remember to turn the survey off when you hit the targeted number of responses to avoid going over budget.

11Export the data from your online survey tool. Update your pre-prepared analysis with the real data. Examine results, and summarize in both text and visual forms. Present the findings to all & sundry. Run a workshop to discuss opportunities and explore the implications. Decide where to focus.

12Stick the research findings on the wall. Share them on your intranet, Google Drive, Slack, or other internal comms channels. And when you hear anecdotes or opinions, remind people we have solid research. Keep referring back to the research through product development.

What did we learn from the process?

Hopefully all this makes Outcome Driven Innovation seem less overwhelming. But as Mies Van De Rohe apparently said, “the devil is in the detail”.

To help you implement ODI more successfully, here’s a quick summary of hard-won lessons from our experience in applying it to Envato Elements:

*Pick an obvious, functional job-to-be-done.

We chose “sourcing a digital asset for a creative project”. Although generic, it was a robust JTBD that would be relevant regardless of which digital assets we ended up focusing on. Plus it also ensured the findings were immediately relevant to our existing business Envato Market.

*Software developers love the logical ODI approach

Developers are often analytical people with a preference for understanding the logic and evidence behind a decision. So they love to see this sort of robust approach that translates in to a prioritized set of customer needs. It removes subjectivity and opinions, and gives them a clear and logical path forward.

*It takes practice to do ODI interviews well.

Drilling down past the initial response into the underlying need is really challenging. People offer up solutions and features, but you need to keep digging to identify the actual outcome they think those things will get them. Bring a colleague along to help out. Prepare templates so you can capture outcomes efficiently. And always do some mock interviews first!

*Do a mock analysis with fake data before you run your survey.

A mock analysis gives you (and those around you) real confidence to know that, once you get the data, analyzing it will be a breeze. This takes execution risk out of the project, and also speeds up the process once you have the data. It also makes your analysis general enough for easy use on future projects.

*ODI surveys are really challenging to take.

Do absolutely everything you can to make the surveys compact and efficient for respondents. With 2 questions (importance & satisfaction) for 100 outcomes, that’s 200 survey questions, plus you’ll probably want some demographic questions too. We used layout, colour, friendly wording, and extensive testing to streamline our survey.

*Give careful thought to survey incentives.

We initially used a $20 Amazon gift card for each participant. Later (on another ODI project for part of Envato Market) we found that a contest & prize pool was cheaper and nearly as effective in motivating participants.

*Once you’ve got the results, make sure you use them!

Keep going back to your results as your product takes shape. Highlight the priority outcomes that you’re aiming to tackle with your new product. One handy way to do this is to display the results prominently in your work space, with emphasis on your target outcomes …and a visual, interesting format! That said, you’ll probably need to keep reminding people in person too.


Are you wondering if ODI helped our new product succeed? Results so far are very promising, with 10,000 paid subscribers in just over a month since launch.


Update: I’ve (finally) set up a Github repository with some of my R code. This includes useful ODI functions, as well as a mock analysis with simulated data. See https://github.com/xrusso/outcome-driven-innovation