Research
Companies and apps
A creator is a company and a company is a creator: both post, both sell, and the one you are studying is usually doing both. Open a company from the advertiser’s name on any ad, from an app under Explore apps, or from an app at the end of a creator’s link, and it comes up as one screen: the apps in the store with an estimate of what they do a month, the ads running from its Meta page, the accounts it posts from, and every link out of its site, named. Watch this company files it in the sidebar under Companies, and the phone shows it under Ideas without the app figures.
Not captured yet
This picture is taken from the running app, not drawn by hand. Run npm run docs:shots to take it.
research-company.png
The tiles
| Tile | Read it as |
|---|---|
| Downloads a month | An estimate, as a low and a high, summed across the company’s apps. The stores publish no download figures; see below for how this one is made |
| Revenue a month | The same, in money. A subscription app’s revenue climbs after its downloads do, and the estimate does too |
| Apps | How many are on file, and on which stores. An app read from one store has its twin on the other found by name |
| Active ads | The ads running from the company’s Meta page in this workspace, and the longest any has run |
| Reach | What Meta says the ads reached in the EU, where it publishes reach. Nowhere else, so this is a floor, never a total |
| Funnel | What kinds of link were found out of their site: community, course, booking, webinar, storefront, app |
An app
Each app has its own panel: the listing’s facts (maker, category, price, in-app prices, rating and how many, release and update dates), the store’s own screenshots in a strip (click one for full size, arrow keys to move along), four tiles, a year of the estimate drawn as a band with the middle as a line (Revenue or Downloads, hover any month for its figure), and Month by month for the table behind the chart with the method printed beside each month. Search terms it ranks for reads Apple’s own suggestions for a term in Apple’s order, how established page one is for each, and where this app sits on it.
How the estimate is made
Apple and Google publish neither downloads nor revenue. Every figure here is modelled from what they do publish, the way every estimate vendor does it, and it is always a low and a high with the method named rather than one confident number. The band on the chart is the honest width of the estimate; it narrows as readings accumulate.
| Method | What it reads | When |
|---|---|---|
| Written reviews | Every App Store review with its date, back thirteen months. A month’s reviews times how many downloads a written review stands for | The first read of any App Store app |
| Rating count | How fast the star count moves between two readings, at how many downloads a new rating stands for | Once two readings a week apart exist |
| Install count | Google’s own install figure, which the page carries exactly, between two readings | Google Play, once two readings exist. This one is not modelled |
| Chart rank | The app’s place on the charts, overall or in its category, on a published curve: the grossing chart for revenue, the free and paid charts for downloads | Whenever the app charts. Replaces the model for this month and scales the history to it |
| Lifetime average | The store’s own counter over the months since release: Google’s install count, or the App Store’s rating count at how many downloads a rating stands for | Before two readings exist, and on the App Store only for an app under three years old, where the average still says something about last year. Weak, and labelled as such |
Revenue is downloads times the share that pay anything in the month, times the representative monthly price from the in-app list, plus the tail of earlier months’ subscribers. A paid app is its price times its downloads. An app that sells nothing inside reads as zero: ads inside apps are not modelled. An app known only from the charts, whose price list has not been read, shows revenue only when the grossing chart places it; opening it reads the list.
Who is posting about it
An app at any size runs creators: a program of accounts named after the app, clipping pages, slideshow pages. Its site never lists them. Organic searches TikTok, Instagram and YouTube for the app’s name in one click and lists the people behind the posts that come back, with their followers, how many posts and their best one; a handle that carries the app’s name is marked runs for the app and sorts first. Each person opens as a profile, one click from being studied. The search lands as a niche in the sidebar like any other, so the posts themselves are there too. Under it, the accounts the site links to.
The funnel
Funnel lists every link out of their site, followed and named: a Skool or Circle community, a course on Kajabi or Teachable, a Calendly, a webinar registration, a Stan store, a Patreon, a newsletter, an application form, an app. Each carries the page’s own title, description and picture, and the button on it. Paste any link of theirs to follow it too; a link-in-bio page is opened and its links listed under it. The same view is on a creator’s profile under Where they send people, seeded with the link in their bio.
Watching, and the phone
Watch this company puts it in the sidebar. Every app of a watched company is read again each night, so the readings the estimate needs accumulate on their own. On the phone, a watched company is under Ideas with its note and funnel. Its apps, prices and figures are read from the app stores, which an iPhone app may not show, so they stay on the web, as do resolving a company and following links.