I often have conversations about whether or not it is worthwhile to spend extra time improving the aesthetics or the “UI” of a feature. I think that some startup people feel like it is less risky to have an app that is beautiful and that is really easy to use than an app that looks really bare bones and has occasional ui hiccups.

This is not the case.

Let’s look at two teams. One team spends that extra time, the other team doesn’t.

The execution mindset: “My great idea should look great”

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

def before_deploy
  take_idea_from_business_plan
  talk_to_designer
  have_meeting
  get_psd
  convert_psd_to_html_by_hand
  design_review
  write_tests
  write_code
  add_ajax
  add_lightboxes
  make_work_in_ie6
  make_work_in_ie7
end

Result: the team builds processes to make it easier to execute. For example, if you insist on using ajax forms in lighboxes throughout the site, then your team makes it really easy to do this using form builders.

The experimentation mindset: “What will people buy?”

1
2
3
4
5
6
7

def before_deploy
  sketch_in_pencil = learning + idea
  code_in_html(sketch_in_pencil_on_paper)
  make_work_in_ie7
  write_tests
  write_code
end

Result: the team builds processes to make it easier to experiment. For example, if you insist on only keeping features that lead to higher revenue per user, then your team makes it really easy to measure revenue per user.

Speed reduces risk

Say the execution mindset takes 100 units of time and money per iteration, and the experimentation mindset takes 10 units per iteration. Say there are two teams that each have 3000 units of time and money before they go bust. Say that each 10 iterations, the team has enough learning to realize that they are on the wrong track and to pivot their product.

The execution team completes 30 iterations, while the experimentation team completes 300. The execution team has pivoted 3 times, while the experimentation team has pivoted 30 times.

The experiment team has had the opportunity to learn 10 times more, and is therefore probably far closer to product market fit.

Less design process leads to faster experimentation. Faster experimentation leads to less risk.