Running a Google Design Sprint


I first came across the concept of Design Sprints whilst helping a UX team with their processes - They were working 2 to 3 months ahead of the development team picking up designs with a total lead time of concept to implementation of up to six months! Just to reiterate this was a lead time of 6 months from idea to first opportunity to test with a customer!

Design Sprints come out of Google and are used widely within Google Ventures - Their Venture Capitalist division. Design sprints are 5 days in duration running  from 10am to 5pm in the following format :-


Set the scene, What problem are we trying to solve? What questions do we need answered? Why are we doing this? Where do we want to be in six months?

Agile retrospective ideas such as 'postcards from the future' are a good way of framing and getting ideas. Asking people to imagine the final result and asking what were the biggest risks that we solved? Or perhaps asking what could of gone wrong is a good way of thinking about experiments to test risky assumptions.  

At the end of day 1 you'll want to know what's the biggest/riskiest question that you want to answer about the product. 

Tuesday - Sketch day

Tuesday is where the team will come up with design ideas, product ideas, etc which will potentially be turned into a prototype.

Spend the morning looking at other products - both in the same field and other fields.. Get people to give demonstrations of ideas they like.. And why they like it.

The afternoon is where you'll start sketching - Individually.... But together! No group think or shouting to stifle creativity! However you'll want to share the space so collaboration can naturally occur.

One final point - These are sketches, On paper! We want them stuck to the wall at some point and spoken about about available for reference not hidden away. 

Wednesday - Decider day

At the start of Wednesday your room should be full of drawings and sketches...

First everyone in turns introduces their design concept, explains and answer questions.

Next people vote on which idea they like best.... There are numerous ways to do this, dot voting is my favourite, give everyone 2 or 3 dots..... Give the key stakeholders 4 or 5 dots..... Yes I know, But they are the key stakeholders!

By Wednesday lunchtime you should have narrowed the choice down to a remaining few. At which point you may decide to combine ideas or possibly to build several prototypes on Thursday and run them on Friday competing.

Wednesday afternoon you'll decide which design to take forward into prototyping, however - despite the temptation.... Don't jump into building it! :-) Another tip... If you decide to go with more than one prototype and you have a couple of high risk ideas consider taking one of these into Thursday. Usually due to the cost/risk high risk ideas never make it off the drawing board however Design Sprints give the opportunity to test with low risk... And sometimes the riskiest ideas are the best.  

Wednesday afternoon is spent storyboarding the solution out - Sketch out the screens designs from a user journey perspective... If you have questions of decisions to be made... Well that's why the main decision maker is there!

Thursday - Build the prototype day

This is where you'll build the prototype... It doesn't need to be real, It can be created in powerpoint or look at some of the web tools that allow you to very easily build prototypes (Marvel) or Sketch - Assuming you're building a IT system and not a physical product. If you are building a physical product look at adapting what you already have or make a flyer or other marketing material for the product.

Each session on the Friday needs to last around an hour (hopefully less) - That will include opening up the interview, setting the scene, demonstration of the prototype (15mins max) and interviewing the customer afterwards as well as writing up any notes etc. 

Friday - Test the prototype.... With real customers! 

This is where existing customers or new potential customers will test the prototype. 

It's important this is done 1-on-1 - No focus groups (They don't work!)

A few words of advice.... Don't use friends, family or even 'tame' users! You're looking for genuine results and feedback.

You need to have around 5 customers booked in for Friday - Any less and it's hard to reach firm conclusions... However you probably don't want too many more either. I once ran a session where we had 7 customers lined up.... And it was probably two, too many!

The output of these sessions may determine the future direction of your product and possibly several months of expensive development... Getting your design decisions right at this stage pays!

If Friday does end in failure - Don't think of it that way... What's actually happened is that 5 days effort has prevented you from wasting months of expensive product development time..... that would have also failed!

Before you run a design sprint... A warning!

It's really important to have the right people involved in these sessions and they must not be considered as part-time sessions where people can dip in and out of.. If you can't set aside 5 days to run the sessions - than perhaps you are not ready to run one!

That all said there will inevitably be some important stakeholders and decision makers who will not be able to put aside 5 consecutive days. 

In which case make sure that the people who really care, those who have a real vested interest in the outcome,  those who are the big decision makers are there for at least the following sessions :-

  • Monday - Preferably all day, they need to set the scene and are ultimately the ones to decide which ideas you'll going to try and test.

  • Wednesday - This is the decider day - Which design will be put to the test? If the 'decider' is not there on Wednesday, If they can't commit to reviewing designs and deciding - DON'T run a design sprint!

  • Friday - You can run Friday without.. But you really want them interviewing and seeing the results of the prototypes - Remember the output of Friday is going to decide your future product development, It's important so treat it as such.

A few other tips....

1. You need a nice large room that you can use all week... You'll be plastering the walls with ideas and sketches - You DON'T WANT to move offices everyday. 

2. You'll going to need whiteboards (lots of white boards), sticky notes (post-its) sharpies, sketching paper, pens, pencils.

3. A facilitator - Preferably somebody inderpendant and somebody who's ran design sprints before! 

4. I've said it before but I'm saying it again... Everybody in the team needs to dedicate time and commit to the process... If the main deciders are not available on Monday and again Wednesday DON'T run the session

5. Don't start until 10am - It gives everyone time to answer emails and do that office admin stuff

6. Have plenty of breaks - You need people to be fresh and alert... And they'll welcome the opportunity to check their phones and those all important work emails. 

7. Take some serious time considering who to invite on Monday and again on Wednesday.... Who has inside knowledge, thoughts, insights that don't reside within your team? Would having somebody from marketing help? Or somebody who works on the support desk and speaks to real customers daily? You may be surprised what others have to offer. 


Design Sprint - Prep notes

Slide deck