Instagram Isn't Testing Read-Only Mode. Your Engagement Numbers Were Already Rented.

Instagram Isn't Testing Read-Only Mode. Your Engagement Numbers Were Already Rented.

On October 3 a reverse engineer named Alessandro Paluzzi posted screenshots of an unreleased Instagram feature called read-only mode. Open somebody's profile in it and you can't like, comment, message, follow, or view stories. By Monday morning the social media trade press was running it as a test. By Monday afternoon an Instagram spokesperson told Engadget that read-only mode "is an internal only exploration, not testing externally at all."

Nobody's testing it. A leaker pulled a string out of an app build and photographed it.

The advice didn't slow. Inside a day the posts hit telling brands to rebuild their reporting, move off likes and follows, and brace for the end of engagement rate. All of it stacked on a feature Meta says it isn't running on a single external account.

What the screenshots actually describe

Read-only mode looks like a perk attached to Instagram Plus, Meta's paid tier. Engadget's own subhead called it a nicer name than stalker mode, and that's the honest read. It's a way for a paying subscriber to study your profile without leaving a trace, and Engadget raised the obvious safety problem with shipping that. A browsing session with no paper trail is useful to a stalker long before it's useful to a shy customer.

It is not a measurement initiative. Nobody in Menlo Park sat down to break your weekly report. The feature that got written up as a threat to your analytics is a privacy toy for subscribers, and it may never leave the building.

The panic was wrong and the direction is right

Everyone who covered this badly got the news wrong. They still got the destination right.

Instagram has spent years building ways to look without leaving a mark. In 2019 it stripped like counts off posts for test users in the US and a handful of other countries. Creators pushed back, because the like count was how they priced themselves to brands, so in May 2021 Instagram handed everybody a switch instead and let you hide the count on any post you choose. The same week the read-only screenshots went around, it started testing a sticker that drops a Threads post into an Instagram Story, which walks your audience into an app where your numbers start from nothing. None of that was aimed at your dashboard. It all landed on your dashboard anyway.

So the lesson doesn't depend on read-only mode shipping. It depends on something that's been true since Instagram turned on its first API. Every number you grade yourself on is a number somebody else can turn off, and Meta has been rearranging how brands get found long enough that this should surprise nobody.

I checked my own scoreboard before I trusted it

This summer I pulled Buffer analytics on 1,460 of my own posts across @gallucciNET, July 26 through September 22. High output, clean cadence, no missed days. The response column was close to empty.

Not a soft quarter. Not a dip I could blame on a holiday. Near zero, across sixty days and four figures' worth of posts.

Read-only mode would not have moved one number in that report. The likes weren't telling me anything before the feature existed, so a feature that hides them takes nothing from me. That's what made me stop using the engagement column as the scoreboard, and it happened months before a leaker found a string in an app build.

Business over that same stretch came in through DMs and LinkedIn. Different column entirely. Meta doesn't run that one.

Sort your numbers by who controls them

There are two kinds of numbers in organic social media reporting, and the difference isn't how useful they are. It's who holds the switch.

The rented ones live inside the app: likes, comments, follower count, reach, impressions, story views, engagement rate, saves. Every one of those is a number Instagram defines, counts, displays, and can redefine or hide on a Tuesday with a changelog note. You don't get a vote and you don't get notice.

The owned ones live where you can see the raw record: sessions on your own site that arrived from social, email addresses on your list, direct messages that turned into a conversation, people who replied to a newsletter, people who walked into the store or showed up at the booth. Those sit in your analytics, your email platform, your inbox, your own count of who came. Meta can make them harder to earn. Meta cannot make them disappear from your records.

Build the report on the owned side. Keep the rented numbers and read them like weather. They're useful for deciding what to shoot this week and worthless as proof the work is paying for itself.

The test that actually tells you something

Forget whether read-only mode ships. Run the free version you've got today. About four minutes.

Open your brand's Instagram on somebody else's phone, logged out. Don't tap a thing. Bio, name field, the link, first nine posts in the grid. That's the whole view a stranger gets. Two questions from there. Does that stranger know what you sell. Does that stranger know how to buy it.

That's the whole exercise. A logged-out stranger who can't interact is already most of your traffic. That shopper existed before anybody leaked a screenshot. On a shelf you'd never accept a package that needed a conversation to explain itself, and the grid is the same piece of real estate with worse lighting.

Most profiles fail it. The bio is a tagline nobody would say out loud. The first nine posts are a seasonal graphic, two reposted compliments, and a photo of the team at a trade show in a city the customer has never been to. Nothing in frame says what the product is or where to get it. It looks fine with the engagement layer attached, because the follower count does the talking. Strip that away and the storefront is a window with the shades down.

What to do this week

Open your last ninety days of organic social media reporting and go through it line by line. Next to every number, write down who can switch it off. The lines where the answer is Instagram, TikTok, or Meta get moved to a second page. What's left on the first page is your report now, and if the first page is thin, that's the finding.

Then fix the bio and the top nine, in that order, and put a real destination behind the link. A profile that reads cold is the cheapest fix available to you, and unlike the engagement rate, nobody can revoke it. That's the work I'd want done before I spent another dollar with a food and beverage marketing agency on content volume.

I keep it to a handful of brands a quarter. If yours should be one of them, message me on LinkedIn.