Both

Kickstart01 · week two · Monday

Testing your solution

Put your prototype or MVP in front of users. Find out whether it delivers value and decide what to improve.

With Alexandru Daicu and Aleodor Tabarcea.

Aleodor

What the two hours are for

You leave tonight with a test you can actually run.

One page. A real person's name on it. A time the two of you have agreed.

Not a reading list.

Aleodor

And then on Friday

Friday, we read what happened.

Not what you hoped would happen. Not the version with the ending tidied up. What happened.

Both

Two of us, because there are two kinds of product on this call

Alexandru Daicu

Product engineer, and a product manager before that.

Mostly B2C. Products where you can put a prototype in front of fifty strangers by Thursday.

Aleodor Tabarcea

Engineering manager, Site Lead at Stripe.

Mostly B2B and enterprise. Products where five users may be all the users there are, and one of them signs the contract.

Some of you can recruit fifty people this week. Some of you will fight for three. We will keep saying which advice is which.

Both

Who is covering what

MONDAY FRIDAY FOUR DAYS IN THE FIELD ALEXANDRU product, B2C ALEODOR engineering, B2B Why we test,and who withThe two questions:value and usabilityObservation beforeinterpretationSaid versus did,and the value verdictThe need, and whatthey already doThe task, the silence,and taking notesThree readingsof one lineCounting usability,and what changes

Monday is preparation, Friday is evidence, and the four days in the middle are the only part of this week we are not in the room for.

Alexandru

Part one

Before the method, the posture

Alexandru

The habit the whole thing rests on

Curiosity over certainty.

You believe in your thing. You have to, or you would not be here on a Monday night.

That same belief is the most expensive thing on this call.

Alexandru

A demo and a test are not the same meeting

A demo

You talk, they nod.

You explain it the moment they hesitate.

You leave feeling good.

A test

They talk, you write.

You watch them hesitate and you say nothing.

You leave with a list you did not want.

Both are useful meetings. Only one of them can surprise you.

Alexandru

The reality check

Your mother would use it.

She would. She would also use a brick, if you built it and sounded excited.

Everybody who likes you will tell you it is good, and not one of them is lying to you.

Alexandru

Our example, across both sessions

Meet Tandem

You say what you are building and what you are missing. It suggests three people a week. You message them.

Tandem does not exist, which is the point. Nobody here has to defend it, so everybody gets to be honest about it.

Alexandru

Two questions, not one

Does this matter to them, and can they work it?

Two questions. Different evidence, different fixes.

Almost everyone only ever tests the second one.

Alexandru

The corner you fear is not the one that hurts

THEY WANT IT THEY DO NOT THEY CANNOT WORK IT THEY CAN WORK IT FIX THE STEPThey want it and they got stuck.A week of work, and you keep your plan.KEEP GOINGThey want it and they can work it.Rare on day four. Enjoy it.THE CHEAP LESSONNobody wanted it, and it barely worked.You spent a week, not a year.THE EXPENSIVE CORNERPolished, usable, and unasked for.Where most of them end up.

Nobody is afraid of shipping something ugly. Everybody should be afraid of shipping something nobody asked for, beautifully.

Alexandru

What we fill in today

One page, seven boxes

  • Who you are testing with, and how you would recognise them.
  • What they do today instead of using your thing.
  • The value question. Does this matter to them?
  • The usability question. Can they work it?
  • The task you will give them.
  • What you will say when they get stuck, and what you will not.
  • The plan. Who, how many, when, where.

The back of the sheet is for what happens while you are testing. That side you bring on Friday.

Alexandru

Box one

Who you are testing with

Alexandru

Who is in the chair?

Everything downstream depends on that one person. Pick the wrong one and you come back on Friday with findings that are detailed, confident and completely worthless.

Alexandru

Describe them by what they have done, not by who they are

What teams write

Aspiring founders, 22 to 35.

Ambitious, early career, tech-savvy.

People who want to start a company.

What you can actually recruit

Has looked for a co-founder in the last three months.

Has had at least one conversation about it that went nowhere.

Is building something now, not someday.

The left column describes a mood. The right one is an instruction you can act on this afternoon.

Alexandru

The reality check

Anyone but the people on this call.

They are the easiest twenty minutes you will ever book, and the least useful ones.

Rehearse on us. Do not count us.

Alexandru

Field sheet, box one

Write who you are testing with

  1. One sentence. Behaviour only.
  2. Something they have done, with a time window on it.
  3. No ages, no adjectives, no "tech-savvy".

If you could not use your sentence to find somebody by Wednesday, it is not finished.

Aleodor

Box two

What they do today instead

Aleodor

Nobody is sitting around waiting for your app.

They are already solving this. Badly, in a spreadsheet, by asking the one person they trust, or by quietly giving up on it.

That is your competition. A spreadsheet and a friend.

Aleodor

Tandem, before Tandem exists

What they do now

Ask the two people they already trust.

Post in a community and wait.

Go to events and hope.

Give up and build alone.

What each one costs them

Trust is high, choice is almost zero.

Free, slow, and mostly silence.

A month of evenings for three conversations.

Nothing, except the company they did not start.

Every one of those is a competitor, including the last one.

Aleodor

The reality check

And if they do nothing at all?

Sometimes an empty space is an opportunity. More often it is just an empty space.

Aleodor

Field sheet, box two

Write what they do today instead

  1. The workaround, in one line.
  2. What it costs them: time, money, or something they gave up.
  3. How you will check that in the test, instead of assuming it.

If the honest answer is "nothing", write that. It is the most useful line on your page.

Aleodor

Break

Five minutes

Alexandru

Boxes three and four

The two questions, written down

Alexandru

The value question

Never ask what they would do. Ask what they did.

"Would you use this?" gets a yes from almost everyone, including people who will never think about you again.

"When did you last go looking for a co-founder?" gets you something that already happened. The past cannot flatter you.

Alexandru

The same curiosity, asked two ways

Questions that flatter you

Would you use this?

Does this sound useful?

How much would you pay for it?

Do you like it?

Questions that cost something to answer

When did you last try to solve this?

What did you do, step by step?

What did that cost you?

What did you do instead, in the end?

If they have to imagine the answer, you are collecting imagination.

Alexandru

The strongest value signal there is

Value costs them something.

A name. A slot in their calendar. A document. Money. A question about when it ships.

"Interesting" is free.

Alexandru

The usability question

Stop asking. Start watching.

Pick the one step you are least sure about, then say nothing at all while somebody tries it. Eight awkward seconds of a stranger clicking around will teach you more than twenty minutes of their opinion.

Alexandru

Tandem, both questions, narrow enough to answer

The value question

Have solo founders actually tried to find a stranger to build with, or do they only ever trust people they have worked with?

Checkable: what did they do, the last time they needed one?

The usability question

Can somebody describe what they are missing clearly enough for a match to mean anything?

Checkable: watch five people fill in the profile step, alone.

Both answerable by Thursday. Neither of them is "is the app good".

Alexandru

Field sheet, boxes three and four

Write your two questions

  1. One value question, answerable from something that already happened.
  2. One usability question, answerable by watching one step.
  3. Next to each, the answer that would make you change your plan.

If no answer would change anything, you are not testing. You are fishing for compliments.

Aleodor

Box five

The task you give them

Aleodor

Give them a goal. Not a tour.

The second you say "now click sign up", the test is over. You have measured whether they can follow directions, and they can, because they are adults.

Aleodor

The same test, two openings

Instructions

"Click sign up, fill in your skills, then look at the matches."

You learn whether they can follow directions.

Every stumble is your fault and tells you nothing.

A goal

"You have decided you need a technical co-founder. See how far you get with this."

You learn where they go first, and what they expected.

Every stumble is a finding.

Say it once. Then the silence does the work.

Aleodor

Field sheet, box five

Write the task

  1. One realistic goal, in their words, not in your feature names.
  2. No steps. No "then".
  3. Read it out loud. If it contains "click", write it again.

Then write down what you say next: nothing at all.

Alexandru

Box six

What you say when they get stuck

Alexandru

Sit on your hands.

Every instinct you have will tell you to help. This is the thirty seconds you came for, and helping is how you delete it.

Alexandru

I'll just show you how it's supposed to work.

All of us, about nine minutes in.

Alexandru

Your whole script, six lines

Six things to say, and nothing else

Silence

Count to eight. Most people solve it, or say something useful, by six.

What are you trying to do?

Brings them back to their own goal without pointing anywhere.

What did you expect?

The gap between what they expected and what is on screen is the finding.

Say that again?

Gets you the sentence word for word, which you will want on Friday.

What would you do on your own here?

Unsticks them without handing over the answer.

Can you show me?

Turns an opinion into a behaviour you can watch.

Alexandru

And three you will badly want to say

Three sentences that end the test

  • "You just need to click there."
  • "Oh, that part is not finished yet."
  • "So what you mean is..."

Say any of them and the rest of the session is just you, talking to yourself.

Alexandru

Field sheet, box six

Practise the silence

  1. In your breakout room, one of you gets stuck on purpose.
  2. The other says nothing for eight seconds. Count them.
  3. Then one prompt from the six. Only one.

Then write down the one you are most likely to blurt out, and cross it off your own sheet.

Aleodor

The back of the sheet

Writing down what happened

Aleodor

The distinction we work on all Friday

Two columns. Keep them apart.

What you saw on the left. What you reckon it meant on the right.

Both are worth having. Mixed together they stop being evidence and turn into a story, and a story is how a team talks itself into the wrong fix.

Aleodor

The same thirty seconds, in two columns

What happened

Scrolled past the profile step twice.

Said: "I do not know what to put here."

Stopped, opened another tab, came back.

Typed "a technical guy, I guess".

What I think it means

The profile step is not readable.

She did not know what we were asking for.

She went to check something.

She does not know what she needs yet.

A camera would have caught everything on the left. The right-hand column is you, guessing, and you are allowed to be wrong there.

Aleodor

Their words, not your summary.

Not the gist. The sentence.

On Friday, one line in their voice will move this group further than a page of yours.

Alexandru

Box seven

The plan for the next two days

Alexandru

How many people

5

For usability. The convention is Jakob Nielsen's, from 2000, and it is about how many people it takes to find the problems in a screen.

It says nothing at all about whether anybody wants the thing.

Alexandru

Two questions, two different recruiting problems

For usability

Five is a reasonable place to stop.

They have to match box one, and that is all.

One at a time, alone, with you watching.

For value

Five is a start, not an answer.

Spread across different situations, not your five friends.

You are looking for a pattern in what they did, not a vote.

Five of the wrong people is worse than three of the right ones, because five feels like enough to stop.

Alexandru

Before tonight ends

Four things on your page before we log off

  • A name. One real person who has said yes.
  • A time, agreed with them, in your calendar.
  • Who runs the session, and who writes.
  • Where the notes live, so both of you can find them on Friday.

A test with no name and no time in it is a plan to test next month.

Alexandru

Field sheet, box seven

Message one person now

  1. Pick the person who most clearly matches box one.
  2. Ask for twenty minutes before Friday.
  3. Say you want to watch them try something, not show them something.

Send it now, while we are all still here and it is equally awkward for everybody.

Both

What changes between now and Friday

Go and be wrong about something.

If you come back on Friday with everything you believed tonight still standing, you did not run a test. You ran a demo with extra steps.

Both

Friday. Bring the back of the sheet.

Messy is fine. Coffee on it is fine. Blank is a problem.

Back to the talks