Ten Competitors and a Name Collision

I had an iOS app for capturing wines you taste at a winery and finding them again months later when you open the bottle at home. It had been sitting untouched since May. I reopened it, wrote objectives for it, and published them.

Then I ran the check I should have run first.

The job

In a user’s words: capture what I am drinking at a winery, and find it again months later when I open the bottle at home.

The rule is that zero or one competitor means the job is open, two or three means contested, and four or more means closed.

Competitor	What it does against this exact job
Vivino	Label photo, then ratings, reviews, notes, food pairings. Scans whole restaurant wine lists
Delectable	Photo identifies the wine in seconds. Own notes, tags people and locations, unlimited free scans
CellarTracker	13.6 million ratings. The established community tasting-note database
Wine-Searcher	Own five-star ratings and notes, saved to a personal cellar
WineLog	Scan or search into a wishlist; tap “I drank this!” and it moves to your logs
WSET Tasting Notes	Label scan plus structured tasting notes
Wine Collection	Label scanning, notes from tastings
Sommo	Camera at a bottle, get region, grapes, style, what to expect
VinoMemo	Positioned explicitly as tasting notes rather than upsell
InVintory	Full cellar management, plus incoming deliveries

Ten, against a threshold of four.

Worse than the count: two of them are built on the specific angles I thought were mine. WineLog ships the capture-then-recall bookend. VinoMemo is positioned as notes rather than commerce. Those were my two differentiators, and both were somebody’s shipped product.

The name

I checked the trip-and-trail angle separately, since that was what my external name pointed at. There is a shipped App Store app called Wine Trails, doing GPS-narrated self-guided winery tours across California and New Zealand. Plus the same name on Google Play, plus a long tail of regional trail apps.

My name was Vine Trails. In store search those are the same string.

So even the surface that would have carried the product was already held, by a competitor, under my name.

The question that actually killed it

Not the competitor count. Recurring intent.

Wine trips happen once or twice a year. That was always the known weakness, and the answer in my own product thinking was that the at-home recall loop is the real engine: trips are rare, but opening a bottle is weekly. Fine reasoning, except that recall at home is precisely what all ten competitors ship. The rare half of the job is unserved because nobody wants it. The frequent half is served ten times over.

Two mistakes, and the second is worse

The first is the obvious one: I wrote objectives for a product before checking whether its job was taken. One search, one paragraph, about a minute. It would have cost less than writing the objectives did.

The second I only found while writing this up. The objectives recorded the baseline as “nothing built, docs only.” That was false. The repository holds 43 source files, a full project, capture-flow UI tests, and a merged rebuild branch through its sixth milestone. I wrote three key results about shipping something that already existed.

That is a different class of error than a mis-set target. It is a set of objectives describing the wrong world. And it exposed a real gap in my own process: I had a rule that every key result must name a measurement instrument, and no rule at all that its baseline had to be read. A baseline is a claim about the world, and it should get checked like one.

The verdict

The interesting part is that the right answer was not “do not build this.”

The app exists. It works. Two people in this household would use it. What it should not have is objectives, a scoreboard, or a store listing, because none of those things make sense for something with ten competitors, annual intent, and a name that collides on the only surface that would have carried it.

So it got re-staged as a personal app. It stays useful, it costs nothing to keep, and it stops consuming attention that was being spent pretending it was a business.

The one thing worth keeping

While counting competitors I searched an adjacent job: tell me what is coming from my wine clubs and when.

Consumer-facing, I found exactly one, and only as a feature inside a cellar-management app rather than as a product. Everything else was software sold to wineries for running their clubs, not to the member receiving shipments from four of them.

That is the same shape as the one thin result in the wine-app sweep and in a car-research sweep before it: every competitor serves the other side of the transaction. It is thin, it is real, and the thing that would serve it is already built and running.

Which is a much better outcome than the one I went looking for.