The Problem

  • We’d had over 100 agencies contact us about our future agency plan, 70% of which would be new customers.
  • From speaking to agencies we knew that our offering wasn't ideal for them because:
    • Agencies wanted to manage styleguides for multiple clients
    • Agencies wanted their work for different clients to be separate
    • Agencies wanted control over billing
    • Agencies wanted the ability to hand over the styleguide to their client with the option to maintain access
    • The pricing of existing features that tackle the above (e.g. guest editors) didn't work for us/them

The 'Uploads' tab, showing how design files are grouped together


  • We wanted to have something working as soon as possible, in order to prevent losing potential agency customers to competitors
  • To help judge demand we added a new CTA to the pricing page of our marketing site. We received 54 submissions through this form between 27th September to 26th January. (For comparison there were approx 250 submissions of the enterprise form in the same time period)
  • Estimated go-live date for an MVP in January.

What does success look like?

  • 10 out of 100 wait-listed teams have bought the agency plan (not on the early-access price)
  • 50% of teams have invited at least 1 client editor
  • 50% of teams are actively using > 1 project
  • 75% of teams say they’ll use zeroheight with their next client

Design sprint

In order to collate the team's ideas and get alignment on a potential solution, I ran a design sprint. I did this because I wanted:

  • A good understanding of the problems, drawing on expertise from the rest of the team
  • Between 1-3 options of product ideas to help solve this problem that we can then validate with potential customers
  • Between 1-2 ideas of pricing + packaging that we can then validate with potential customers

The design sprint took a total of five days;

  • Day 1: Map out the problem and focus
  • Day 2: Sketch solutions
  • Day 3: Decide which idea to prototype
  • Day 4: Create a high-fidelity prototype
  • Day 5: Test the prototype with users

Customers and key players on the left and our completed goal on the right

Some of my sketches from the design sprint


Our chosen solution, ready to be taken to high-fidelity for testing


In order to validate or disprove any assumptions I tested an initial prototype with 7 users who had expressed an interest in a zeroheight agency plan.

The template script I created for interviewing users who'd expressed interest in an agency plan


The prototype I tested with participants - see the full prototype here

Some of the key learnings that came from the interviews

Simplifying our MVP

After affinity mapping and prioritising the feedback from our interviews, we decided to trim down our concept to something we felt offers the most value at the smallest effort to build. This included, but was not limited to;

  • Removing the project switching panel on the left-hand-side in favour of a simplified dashboard
  • Removing the contextual team manager panel on the right-hand-side
  • Simplifying the dashboard project cards to only display styleguides
  • Only supporting Figma (75% of our users) to start with
  • No support with our Slack Integration
  • No support from syncing pages between projects
  • No support for users to duplicate or transfer styleguides or design uploads between projects

The simplified project dashboard

After creating a project users can invite the whole agency team and/or someone from the client side

Within a project users can manage styleguides, uploads and settings specific to that project

How's it going?

16 out of 100 wait-listed teams have bought the agency plan, not on the early-access price 🚀
Our target was 10

0% of teams have invited at least 1 client editor 😥
Our target was 50% but the initial build takes time

25% of teams are actively using > 1 project ⏰
Our target was 50%, again this might be because their efforts are on a single project for the time being

75% of teams say they’ll use zeroheight with their next client 💬
We don't know yet, but are sending out an email questionnaire to find out