Escaping the discovery doom loop

Who are we?

@ames.world

@priyanca.bsky.social

Working on services in government

PD

Working on services in government

Expectation

Reality

Why we’re doing this talk

PD

Who is this for?

IA

What is a discovery doom loop?

When a service can’t get itself out of the discovery phase.

Frustration

Confidence is lost

Time and money wasted

Command and Control creeps in

The opportunity to explore is gone.

Causes of discovery doom loops

Doom Loop 1:

​

Unclear problem to solve

Do a discovery

“We’re not clear on the problem”

We unearth multiple problems

​

Doom Loop 2:

​

Predetermined solution

Go and find out the user needs

“We’ve decided this solution”

The needs don’t align with the solution

​

Doom Loop 3:

​

False Certainty

Can only discover so much in a short period

“We need certainty”

We need more certainty

​

You can have combinations of loops at the same time

You can have combinations of loops at the same time

You can have combinations of loops at the same time

Each of these scenarios require slightly different tactics

But you can use these tactics together if you have multiple loops

Doom Loop 1: Unclear problem to solve

PD

Symptoms

A brief that makes no sense to anyone outside of the project.

Differing focus when talking to stakeholders individually.

Team can’t clearly express what’s expected of them.

Problem not described in terms of impact on the end user.

Do a discovery

“We’re not clear on the problem”

We unearth multiple problems

​

Reminder, what does a good discovery brief look like?

https://www.slideshare.net/slideshow/the-actual-problems-to-be-solved/73017206#1

Reminder, what does a good discovery brief look like?

https://www.slideshare.net/slideshow/the-actual-problems-to-be-solved/73017206#1

Tactic 1: User research your stakeholders

Tactic 1: User research your stakeholders

1. Interview stakeholders
2. Identify themes, patterns, points of contention?
3. Can you write a more coherent problem statement?
4. Bring the stakeholders back together to playback
5. Agree the new scope
6. Be explicit about anything that’s been descoped

Tactic 2: Mini discovery

IA

Tactic 2: Mini discovery

• Your stakeholders struggle to create a brief, it may be necessary to do a short piece of discovery activity to help frame the problem.
• If so be explicit about that, we are doing some xxx to help understand the problem space

​

Tactic 3: Bring your stakeholders with you

Tactic 3: Bring your stakeholders with you

• Engage and involve stakeholders in research cycle, deepen their understanding of what UCD people do and what we are learning in real time
• Invite them to show and tells, encourage them to feedback.
• Take every opportunity to repeat the problem statement, put it at the beginning of all show and tells and weeknotes or other comms.
• Depending on the size and scale of your show and tells, it may be necessary to give them another avenue to feedback.
• Make them feel more comfortable with uncertainty.

It’s ok to feel like this

Doom Loop 2:

The predetermined solution

Symptoms

The brief mentions the solution or product.

The policy determines the end product.

Tactic 1: Due diligence

Tactic 1: Due diligence

1. Who are the users and what are their needs?
2. What parts of the service journey will the solution support?
3. What are we planning to do about any gaps?
4. How much will it cost?
5. Is it worth it?

Tactic 2: Prototype

Tactic 2: Prototype

Alistair Ruff: https://www.etsy.com/uk/listing/1761872459/service-design-prints-a6-postcard

“Starting to build things surfaces assumptions and misalignment. It helps communicate and build momentum. It can feel wrong to start building early, but it’s a very effective way to learn.”

Tactic 2: Prototype

​

1. Be careful not to get bogged down in full build
2. Speak to suppliers about demo versions you can use.
3. If that’s not possible ask the supplier to give you a service map or user journey of how people complete a task in their software.
4. If you can’t get either of those things, that’s an early warning that the solution may not be right.

​

Tactic 3: Motivating the team

PD

Tactic 3: Motivating the team

• Be realistic
• Acknowledge constraints and context
• Do the work to a high standard anyway
• Recognising and calling out wins

Doom Loop 3:

False certainty

IA

Symptoms

Massive scope.

Lots of different user groups.

Complex problem space.

Endless cycles of discovery without delivering anything.

​

Can only discover so much in a short period

“We need certainty”

We need more certainty

​

Tactic 1 - Look for points of similarity

PD

Tactic 1: Look for points of similarity

Your problem is likely not as unique as you think it is.

Are there existing services that bear similarities to the problem you’re trying to solve?

Can you reduce uncertainty by translating and applying what you know about similar services?

​

​

​

Tactic 2 : Breakdown scope

IA

Break down scope - products

Jamie Arnold / Public Digital

Tactic 3: Always be discovering

PD

Tactic 4: Always be discovering

• The discovery phase won't be the only opportunity you'll get to do exploratory research
• You don't need to cover absolutely everything
• Don't put too much pressure on yourself
• You need just enough information to get you to the next point
• Discovery is important but you can still change things in subsequent phases

Tactic 4: Always be discovering

Tactic 4: Talk about cost

IA

Tactic 2: Talk about cost

Conversations to have:

• Timebox
• This discovery will cost x, how does that compare to the expected service (ideally) or project costs?
• This week will cost, what do we need to learn?
• This week, service failure cost x, what can we do now to reduce this now?
• Cost of delay, not solving this problem by x will cost y.

Tactic 5: Front-load delivery capability

Tactic 3: Front-load delivery capability

​

Add some devs to the team early

Prototype some ideas

Make small changes to Live

Production probes to test ideas

Radiate intent

Tactic 5: Stop

Tactic 5: Stop

​

​

Ask yourself:

“Are the requests for more certainty because we’re not being brave enough to stop?”

​

Tactic 5: Stop

Be bold, make the case for stopping.

Talk about costs (so far and if we continue).

Talk about other priorities.

Talk about user needs (or lack of).

Drive for a decision to continue to build, or to stop.

It may be necessary to stand down the team.

This may take as long as the actual discovery.

In Summary

There are many potential doom loops in discoveries

Doom Loop 1: Unclear problem to solve

• User research your stakeholders
• Mini discovery
• Bring your stakeholders with you

Doom Loop 2: The predetermined solution

• Use due diligence as a framing
• Prototype!
• Motivate the team

​

Doom Loop 3: False certainty

• Look for points of similarity
• Break down scope
• Always be discovering
• Talk about cost
• Front Load delivery capability
• Stop!

It’ll be ok!

PD

Thank you!