When nobody knows you, polish is the only evidence

A few days ago Luis von Ahn, the CEO of Duolingo, wrote something that put words to something I had been doing without quite knowing why.

His point was that Duolingo does not ship MVPs. They ship V1s, and the two are not the same thing at all.

An MVP exists to test whether an idea is worth pursuing. It is meant to be unfinished, so rough edges come with the territory. A V1 can be small, it can be missing most of what you eventually plan to build, and it still has to be finished. Think of it like having good copy, proper design, everything that you have planned functioning well.

What stayed with me from his post was the following thing. He says MVPs give people permission to ship sub-par work, because "it's an MVP" is an excuse that always works. He ends with something close to: if it is not good enough to put your name on, then don’t ship it yet.

I read that and recognised an argument I had been having with myself for weeks.

Why this matters more when you are small

Duolingo can ship something rough and survive it. This happens as hundreds of millions of people already know what the company is and they really value their product and what the company stands for. So one disappointing feature re-evaluate the whole product because of it.

A fairly new company which is actively building their reputation in that very moment has none of that. When somebody arrives on my site, whatever they see is the entire dataset. For example, where there is no track record to support your claims and no reputation absorbing fair mistakes that any company may have at first, the people coming across your company first conclusion that your product or your competency itself is half-finished instead of fully rounded and polished.

Which is the part I think gets missed in the usual MVP advice. For a company with a brand, shipping rough is a quality decision. For anyone without one, it is a trust decision, and those are not the same calculation at all.

I wrote a few weeks ago that every business relationship runs on trust and that you either build it slowly or spend what you already have trying to push it fast. This is the same idea seen from another angle. When you have none of either, the work itself is the only evidence available. Polish is not vanity or hesitation in that situation.

Where I still use an MVP

I want to be careful not to turn this into a rule I do not actually follow, because I do ship MVPs. Just not as products.

When I want to know whether an idea is worth something, I build a landing page and I try to collect interest people may have in that product or idea but gathering a minimum of an email address.

That is an MVP in every meaningful sense, and rough is completely fine there, because the aim at that point is sell a half-baked product but to judge whether there is a market or interest in that product at all.

So the rule I actually work to is narrower than "always polish." Validate rough. Ship finished. The landing page, copy, harmless email automation flows are where you are allowed to be scrappy. The thing people actually use and should create value for them is where we all need to be more careful.

What I nearly shipped

Here is where this stopped being theory for me.

When I put the new offering live, I needed a way for people to find out whether their own site has a visibility problem. The obvious version, the one I could have shipped in an evening, was a page explaining the problem with a button saying book a call and I will tell you what is wrong with your site.

That is a perfectly reasonable MVP. It tests the idea. It costs almost nothing.

It also only works on people who already trust me. To get anything out of it you have to hand over half an hour of your time to a stranger, on the strength of a page that has told you nothing specific about your own situation yet. The value sits entirely on my side of the conversation, and the visitor has to move first to reach it.

So I built the other thing instead. You give it your website address and it scans the site the way a search engine and an AI assistant would, and it shows you what they see. Which pages are reachable, what is missing, where the gaps are. You get the answer yourself, on the page, without talking to me.

It took considerably longer than an evening. And the trade is obvious once you see it written down: you get the value before you have to trust me with anything. If you look at the results and decide you want help closing those gaps, that is a good outcome for me. If you look at them and fix everything yourself, that is fine too, and you now know something useful about your business that you did not know before that morning.

Go and run it on your own site

It is free, and it takes about a minute.

→ Scan your site with Scout

Put your address in and see what the engines actually see. Most people have never looked, and it is rarely what they expect.

If something comes back that you cannot make sense of, reply and tell me what you are seeing. I read every reply and I would rather have that conversation than any other kind.

What I am still unsure about

The first is that I do not yet know whether this was the right call commercially. It is early. The polished version took longer to build, and a rougher version would have been live sooner. I personally believe the trade was correct but only the time and the market will tell.

The second is that the line between small and unfinished is genuinely hard to see from the inside. Everything I build feels unfinished to me, because I know what I planned to build. The only test I have found that works is the one von Ahn describes: would I be comfortable with my name on this in front of somebody whose opinion I care about. That question is uncomfortable and it is usually right.

What I'm Thinking About This Week

Whether the MVP idea has quietly become permission rather than a method. It was meant to be a way to learn something cheaply before committing. Somewhere along the line it turned into a reason not to finish things, and the word is now doing work it was never supposed to do. I have used it that way myself, and the honest version was usually that I was tired of the project.

Reply to this email with one of:

• What is the last thing you shipped that you would not have put your name on?

• Do you validate cheaply and ship properly, or have the two blurred together?

• Run Scout and tell me the one thing in the results that surprised you.

P.S. The toolkit, in the order I would use it: Scout → Visibility Audit → GEO/AEO Audit Checklist → How to Get Cited by AI → SEO Strategy Playbook → Title & Meta Audit → Sovereign Idea Workflow. All free.

P.P.S. Know someone sitting on something they keep calling an MVP? Forward this. The question at the end is the useful part. They can subscribe at thesovereigntechnologist.com.

That’s all for this week.

See you next Thursday.