What Instagram actually returns for an architecture firm: numbers from the accounts we run

Almost every article about Instagram for architects is advice without evidence. Post consistently, use reels, tell a story. Probably true, and impossible to check. What is missing is the boring part: what the numbers actually look like when someone runs these accounts for a living.

So here are ours. I am the co-founder and CTO at Skala Social, I am an engineer by training, and these are figures from architecture and design accounts we manage, taken from the monthly reports we send those clients. Where a client is named, it is because they agreed. Where they are not named, the figures are still theirs.

What this is based on

This is not a survey of the industry. It is a small number of accounts we run directly, which is the trade: fewer data points, but real ones, measured the same way every month.

• A Berlin architecture practice (Sterling Presser Architects + Engineers), around 1,500 followers at the start. Window: a normalised monthly average across two and a half months.
• An architecture educator in the Middle East selling online courses, small account, launching to an audience that did not exist yet. Window: 30 days.
• An architectural visualisation studio in Dubai, the largest of the group.
• A master's programme in Madrid (MCH Collective House).

Two of those come with a full month-by-month funnel, which is what the detailed numbers below are drawn from. Account sizes are small to mid, which is deliberate. Most architecture practices are not starting from 100,000 followers, and benchmarks from accounts that are will only mislead you.

The funnel we actually measure

Followers are the number everyone quotes and the least useful one. We track four steps, in this order, because each one fails differently:

1. Reach. How many accounts saw the work at all.
2. Interactions. Likes, comments, saves and shares. Measured against reach, not against followers.
3. Profile visits. How many cared enough to look at who you are.
4. Link clicks and inquiries. How many left Instagram for your site, or wrote.

A drop between any two steps points at a different problem. Weak reach is content and distribution. Strong reach with weak profile visits is a hook problem: people watched and did not care who made it. Strong profile visits with no clicks is a profile problem, which is the cheapest thing on this list to fix. We wrote up how to read this in Instagram analytics for architects.

The Berlin practice, in numbers

A working architecture office, no existing audience to speak of, no advertising budget. Normalised monthly averages, before our work and after:

• Accounts reached: 955, then 11,000
• Views: 6,510, then 38,400
• Interactions: 68, then 1,072
• Interaction rate: 7.1%, then 9.7%
• Profile visits: 657 after (not tracked before)
• Clicks to the site: 2, then 16
• Followers: 1,501, then 1,750

The follower number is the least impressive one in that list, and that is the point. Followers grew about 17%. Reach grew roughly twelvefold and interactions roughly sixteenfold. An architecture practice does not need a large audience, it needs the right few hundred people to see the work and then click through. Sixteen qualified visits to your website a month is a different business than two.

The educator account, in numbers

Different situation: an architecture educator launching online courses, starting essentially from zero attention. Over thirty days:

• 200,000 accounts reached, 347,000 views
• 44,000 interactions, an interaction rate of 21.9%
• 10,000 saves and 8,500 shares, which is 23% and 19% of all interactions
• +2,663 followers, a 63% increase in a month

The interaction rate is the figure worth pausing on. Nearly 22% of the accounts reached did something, and almost half of those interactions were saves or shares rather than likes. Saves and shares are the two signals that carry content beyond its existing audience, which is exactly why that account grew as fast as it did.

The other two accounts

The Dubai visualisation studio and the Madrid master's programme were not tracked on the same full funnel, but both point the same way.

The Dubai studio had good visuals and had been buying ads that brought views rather than inquiries. Moving to organic content produced a fivefold increase in conversion on 100% organic traffic, with no ad budget at all, and around 600 new followers alongside it. Reach was never the problem there. Relevance was.

The Madrid master's programme needed applicants rather than an audience: roughly tripled reach, +30,000 views and +400 followers, with the point being which people those views came from.

The finding that surprised us

Put the two accounts side by side on one ratio, profile visits as a share of accounts reached:

• The large account: 3,800 profile visits from 200,000 reached, or 1.9%
• The Berlin practice: 657 profile visits from 11,000 reached, or 6.0%

The smaller account converted attention into profile visits roughly three times better.

That runs against the instinct that bigger reach is better. Broad reach on a viral piece brings a lot of people who are interested in the video, not in you. A tighter, more relevant audience brings fewer people who are far more likely to want to know who is behind the work. For a practice looking for clients rather than an audience, the second is worth more per view.

The practical version: if your reach jumped but nothing downstream moved, you did not have a bad month, you reached the wrong people.

What actually performed

Across these accounts the two best-performing pieces were not renders of finished buildings.

The strongest single post was a carousel: "Everything I learned as an architect in 30 years, in 10 slides." 142,000 views, 21,000 interactions, 9,800 likes, 5,500 saves. Experience, compressed, in a format people save to read again.

The strongest reel was a course announcement built as a teaching piece rather than an advert, at 41,000 views and 4,500 interactions, with a save rate near 20%.

The other pattern that reliably worked was a trend-riding post, an animated render together with an explanation of how it was built, published on a sound that was travelling at the time. That one took analysis rather than luck: we looked at what was moving in the format first, then made something honest that fitted it.

The thread running through all of them: a person behind the object. How this architect got here. What problem they solved and how. People buy from people, and a prospective client is not looking for someone who draws a nice building, they are looking for someone who will solve their problem. A render proves the first. A story proves the second. That is also why a deliberately debatable topic works: people argue in the comments, and the argument carries the post.

Cadence: regularity beats volume

We publish sixteen posts a month on Instagram as standard, fewer where a client prefers it. The exact number matters less than the fact that it does not stop.

The failure mode we see constantly is forty posts in one week followed by a year of silence. That does not build anything. A practice posting steadily at half the volume will beat it comfortably, because the account is never rebuilding from cold, and because the people who found you last month are still being reminded that you exist.

Follower growth across the accounts we run tends to sit around 10% a month, and reach typically passes whatever the account was doing before within about three months. Profile visits and outbound clicks tend to move by multiples rather than percentages once the profile itself has been rebuilt to lead somewhere, which is usually the single cheapest improvement available.

What we cannot promise

We cannot promise these numbers, and neither can anyone else. This is a small set of accounts, over short windows, in specific situations. An educator launching a course and a working architecture practice are not the same business and do not behave the same way. Instagram changes what it favours, sometimes abruptly.

What is repeatable is the method, not the multiplier: measure the four steps separately, fix the one that is actually broken, publish at a rate you can sustain, and put a person in front of the work. What that produces on your account is a question we would answer honestly rather than with someone else's numbers.

Common questions

What is a good engagement rate for an architecture account? Measured against reach rather than followers, the accounts here ran between 7% and 22%. The lower end is a healthy working practice, the upper end is content being saved and shared heavily. If you are measuring against follower count instead, the number will look very different and mean less.

How often should an architecture firm post on Instagram? We run sixteen posts a month as standard. Consistency matters more than the exact count. A sustainable rhythm beats a burst every time.

How many followers does an architecture firm need to get clients? Fewer than you would think. The Berlin practice above went from 2 to 16 site visits a month at under 2,000 followers. Relevance of reach matters more than size of audience.

Do reels work better than carousels for architects? Both, differently. The single strongest post in this data was a carousel, because the format suits compressed expertise that people save. Reels did the reaching. Use reels to be found and carousels to be kept.

How long before an Instagram account shows results? Reach tends to move within weeks. In our experience an account passes its own previous performance within roughly three months. Inquiries follow the profile and the funnel, not the follower count.

The honest summary

Instagram for an architecture practice is not a popularity contest, it is a small funnel with four measurable steps. Reach is easy to buy attention for and easy to waste. The valuable part sits further down, where a few hundred relevant people decide whether to look at who you are, and then whether to click.

If you would rather hand that to people from your own field, tell us about your firm. We run Instagram and LinkedIn for architecture and engineering firms, and the monthly report we send is where every number in this article came from.