Build in public in 2026: what still works and what doesn't
26% of new indie product sites link X and 2% link Threads. An evidence-led guide to building in public in 2026: reach, links, formats, risks and a routine.
14 min read
Key takeaways
- Of 1,418 indie products that passed our review since mid-September 2026, 26% link an X account from their own website; 2.0% link Threads and 1.7% link Bluesky.
- Meta says Threads passed 500 million monthly actives by July 2026 and has since overtaken X. A SpaceX filing put X and Grok together at about 550 million monthly users in March 2026.
- X's published ranking code weights a predicted reply at 5.0, a like at 0.708 and opening a link at 0.2. On original posts from mutual follows, the reply weight rises to 20.
- In Buffer's 2026 format data, link posts had the lowest median engagement rate on X (2.25%) and Threads (2.34%), while LinkedIn document carousels led every format at 21.77%.
- Baremetrics' Open Startups page listed 20 companies sharing live revenue in 2018 and lists 5 today. TrustMRR says it lists over 15,000 startups with provider-verified revenue.
Building in public still works in 2026, but the version that works is narrower than the one that made the hashtag famous. The feeds that carry maker updates now reward conversation and time spent on the post itself. They do very little for a bare link, a revenue screenshot or a daily status line. X publishes the code that ranks its For You feed, and in it a predicted reply carries a weight of 5.0, a predicted like 0.708, and opening a link 0.2 (X, x-algorithm).
The audience has moved more than the makers have. Meta said in July that Threads had crossed 500 million monthly actives, and in October that it had overtaken X (Meta, Q2 2026 call; Meta, Q3 2026 call). Yet of 1,418 indie products that passed review on Awesome Indie since mid-September, 26% link an X account from their own website and 2.0% link Threads.
The open-startup idea has changed too. Publishing revenue used to be a statement of values; now it is mostly proof. Baremetrics' Open Startups page, once a roll call of companies sharing live revenue, lists five (Baremetrics). TrustMRR, which reads revenue directly from payment providers, says it lists more than 15,000 startups (TrustMRR).
Where is the build-in-public audience in 2026?
Every platform counts differently, so this is not a league table: these are the latest figures each company, or a securities filing, has put on record.
| Platform | Latest figure | What is counted | As of |
|---|---|---|---|
| More than 1.4 billion | Members (accounts, not active users) | October 2026 (Microsoft) | |
| X (with Grok) | About 550 million | Monthly active users of X and Grok combined | March 31, 2026 (SpaceX S-1) |
| Threads | More than 500 million, and "overtaken X" | Monthly actives | July and October 2026 (Meta) |
| 130.3 million daily, 515.2 million weekly | Daily and weekly active uniques | Q3 2026 (Reddit) | |
| Bluesky | Over 43 million | Users, as Bluesky counts them (accounts) | March 2026 (Bluesky) |
The X figure bundles Grok in, and account counts flatter Bluesky: Similarweb estimated its mobile app at 10.4 million monthly active users in June 2026, down 27.2% year over year, while X's mobile daily actives fell 7% to 123.7 million in July (TechCrunch).
Where indie makers point their visitors
We looked at which social accounts the websites of new indie products link to. The sample is every product submitted to Awesome Indie between September 16 and November 12, 2026 that passed review and whose homepage our system could read: 1,418 products.
- X26%
- LinkedIn20%
- YouTube14.8%
- Instagram13.7%
- Facebook9.9%
- GitHub9.4%
- TikTok4.5%
- Pinterest4.1%
- Threads2%
- Bluesky1.7%
1,418 products submitted September 16 to November 12, 2026 that passed review and whose homepage we could read. A product can link several networks.
Show the numbers
| X | 26% |
|---|---|
| 20% | |
| YouTube | 14.8% |
| 13.7% | |
| 9.9% | |
| GitHub | 9.4% |
| TikTok | 4.5% |
| 4.1% | |
| Threads | 2% |
| Bluesky | 1.7% |
X is still the default identity for indie makers: 368 sites link an X account, about 13 times as many as link Threads (29) and 15 times as many as link Bluesky (24). Bluesky is an addition, not a replacement: 17 of the 24 sites that link Bluesky also link X. And most products link nothing at all: 59.3% of the sample links none of the ten networks, which matters later when we get to launches.
Our data cannot say whether those accounts are active, or whether the gap between where makers are (X) and where Meta says users are (Threads) is an opening or a sign that buyers of indie software still sit on X. For what these products are, see what indie makers are building in 2026.
Do posts with links still lose reach?
On X and Threads, yes, by the numbers. The platforms now explain it as an engagement problem rather than a penalty, and the result for a maker is the same.
X. Elon Musk said it plainly in 2023: "Our algorithm tries to optimize time spent on X, so links don't get as much attention" (TechCrunch). In October 2025, head of product Nikita Bier gave a different reason: link posts get less reach because "the web browser covers the post and people forget to Like or Reply," and X tested an iOS browser that keeps the post visible (Engadget). He denied that links are "deboosted" (Nieman Lab). The side effect: the new browser preloaded pages before anyone tapped, and sites including Substack and Bluesky saw inflated, partly fake traffic from X (The Verge). In Buffer's study of 18.8 million X posts, link posts from accounts without Premium had a 0% engagement rate as of August 2025, against about 0.40% for text posts (Buffer).
Threads. In May 2025 Meta allowed five links in a bio and said posts with links would be surfaced more often in recommendations (Engadget). Adam Mosseri later said "Links have been working much better for more than a month now," but Chartbeat data showed Threads still sent less than a tenth of a percent of publishers' referral traffic (Engadget).
LinkedIn. LinkedIn's write-up of its rebuilt feed (March 2026) says ranking reads a member's sequence of actions, "long dwells, likes, comments, shares," and it does not mention links at all (LinkedIn Engineering). We found no LinkedIn documentation that says link posts are ranked lower, or that they are not. What LinkedIn does report, in its July 2026 business update, is that time spent reading comments rose 18% (LinkedIn).
Bluesky. Feeds are a marketplace: "anyone can create a feed," and users choose (Bluesky). In October 2026 it added profile links at the top of a profile (Bluesky).
Buffer's 2026 format analysis puts numbers on all four, as medians across its customers' accounts (Buffer):
| Platform | Best format | Link posts | Worst format |
|---|---|---|---|
| X | Text, 3.56% | 2.25% | Links |
| Threads | Video, 5.55% | 2.34% | Links |
| Document carousels, 21.77% | 3.81% | Text, 3.18% | |
| Bluesky | Video, median 5 interactions | 3 interactions | Links and text, tied |
Read it as direction, not law: these are Buffer customers, Bluesky is measured in interactions, and the leaders change between editions. The direction is consistent, though: put the substance in the post and treat the link as the second thing. When you do share one, the preview image is most of what people see, so check what yours renders with the Open Graph checker.
What X's own ranking code rewards
X publishes its For You ranking code on GitHub. A model predicts how likely each viewer is to take each action on a post, and the final score is a weighted sum of those predictions (X, x-algorithm). The current weights include (X, param.rs):
- Like 0.708, repost 1.0, follow the author 4.0
- Reply 5.0, quote 5.0, share by DM 5.0, share by copying the link 20.0
- Open a link 0.2
- "Not interested" -53.07, block the author -31.2
X warns these are not exchange rates: they multiply predicted probabilities, and a rare action needs a large weight to move anything ("one report cancels 468 likes" is, in its words, incorrect). Still, the design values conversation and sharing, not clicks away.
Two 2026 changes sharpen that. In July, X tested and then rolled out a boost for original posts from people you follow who follow you back: the reply weight on those posts goes up by 15, so 20 instead of 5 (X, bidirectional follow boost). And in October, X updated the ranking model to predict time spent better, and retrieval to surface "posts that people read all the way through" (X, x-algorithm). The code also discounts posts from accounts a viewer does not follow.
For a maker, that turns into three rules. Your mutual follows are your distribution, so build them by replying to peers, not by posting more. Write posts that are worth finishing. And a post that makes people tap "not interested" costs far more than a dull one.
Replying to your own comments also lines up with better results. Buffer found posts where the author replied had 5 to 42% higher engagement, depending on the network (42% on Threads, 30% on LinkedIn, 8% on X), though it cautions this is correlation, not proof (Buffer).
Is sharing revenue still worth it?
Buffer is the longest-running precedent. It has published its finances and salaries since 2013, when it was 17 people at $2.2 million in annual run rate (Buffer); its public dashboard showed $28.4 million ARR when we checked in November 2026 (Buffer). Its founder is candid about the cost: transparency "does inherently create extra work for us," and before going public the team worried competitors would find it easier to poach staff.
The movement Buffer helped start looks different now. We counted the companies on Baremetrics' Open Startups page, where startups show their live monthly revenue, across archived snapshots:
- June 201820
- January 202023
- January 202219
- January 20246
- November 20265
Counted from Internet Archive snapshots and the live page. One vendor's showcase, not the whole open-startup movement.
Show the numbers
| June 2018 | 20 |
|---|---|
| January 2020 | 23 |
| January 2022 | 19 |
| January 2024 | 6 |
| November 2026 | 5 |
Revenue sharing did not die. It moved to verification. TrustMRR connects to a founder's payment provider (read-only for most providers) and says it lists more than 15,000 startups, over 3,000 of them for sale (TrustMRR). It also warns that its figures can differ from a provider's dashboard by up to 30%, and its anonymous mode hides a startup's name but keeps the metrics public (TrustMRR FAQ). The reason to publish revenue in 2026 is mostly to prove a business is real, often on the way to selling it. If that is your reason, a verified page beats a screenshot. If it is not, you lose little by keeping the number private.
The research on transparency supports showing the work more than showing the money. In five experiments, Buell and Norton found people valued a service more when it showed the effort behind it, because perceived effort made them feel reciprocity (Management Science). It is a 2011 study of online services, not social posts, so treat it as suggestive. It backs the part of building in public that costs nothing to give away.
The real risks, with examples
Copycats. Cory Zue, who makes SaaS Pegasus, wrote that both of his main products "have been copied many, many times," one with the "same UX, same pricing, same sections on the landing page" (Cory Zue). Mike Rubini, who had shared his metrics and launch strategies in detail, found competitors making "comparison pages all about me," one still using his "exact strategies from a year ago"; after he stopped sharing, he wrote, "my competitors didn't grow as fast as me, because they lost their source" (FlatNine). Arvid Kahl wrote about an earlier case of Rubini's: his product Groouply was cloned, and the cloner put up a comparison page trashing the original (The Bootstrapped Founder). These accounts date from 2021 to 2023.
Revenue as bait. Zue describes the reader who thinks: "whoa, he's making $10k a month selling that? I could build that in a weekend."
Fatigue. Zue published 123 retrospectives over six years before concluding they had become "a crutch" and that he had "stagnated." Buffer, with a whole team, still calls transparency extra work.
The counterweight is also in Zue's essay: most copycats "flail around a bit" and give up. What a clone cannot copy is the relationship you have with the people following along, which is the thing building in public is for.
A weekly cadence a solo maker can keep
About three hours a week, on one primary network, plus one other at most. Choose the network where your customers are, not where other makers are.
- Monday, 30 minutes. One post about a decision you are making this week: what you chose, what you rejected, why. Decisions invite replies; status updates do not.
- Tuesday to Friday, 15 minutes a day. Reply thoughtfully to 5 to 10 posts from people in your niche or your customers. On X, that is how mutual follows (and the reply boost) are built.
- Wednesday, 30 minutes. One show-the-work post: a screenshot, a 20-second screen recording, a before and after. On LinkedIn, make it a short document carousel.
- Friday, 30 minutes. One number with its lesson: "the new onboarding step moved trial-to-paid from 4% to 6%, here is what changed." Ratios teach; absolute revenue is optional.
- Same day, every time. Answer the replies on your own posts.
- Once a month, 90 minutes. A longer write-up on your own site (a changelog, a teardown, a post-mortem) that outlives the feed. Share it in a post that stands on its own, link second.
On Reddit, participation comes before promotion by rule: "Participate authentically in communities where you have a personal interest, and do not spam" (Reddit Rules). Read each subreddit's own rules before posting your product.
What to share and what to keep
| Share | Keep | Why |
|---|---|---|
| Decisions and the reasoning behind them | The exact acquisition channel that is working right now | Reasoning builds trust; a live playbook is what copycats lift (Rubini) |
| Screenshots, short demos, before and after | Unreleased pricing and landing page structure | Clones copy "same pricing, same sections" (Zue) |
| Customer problems in their words, anonymized | Customer names without permission | Consent, and your customers are not your content |
| Mistakes and how you fixed them | Revenue, unless you have a reason to prove it | Revenue recruits competitors; mistakes earn sympathy |
| Ratios and experiment results | Unverified revenue screenshots | A screenshot proves nothing in 2026; verified pages exist |
How to tell if it is working
Followers and impressions are the wrong scoreboard. Measure what building in public is supposed to produce: people who try and pay for your product.
- Tag every link you post with UTM parameters (
utm_source,utm_medium,utm_campaign) so your analytics can attribute visits (Google). - Count signups and paid conversions by source, not visits. Visits are easy to inflate, as X's preloading episode showed.
- Ask new users where they heard about you in a free-text field. People remember "saw you on X" even when they never clicked.
- Track one soft signal weekly: replies or messages from potential customers, not other makers.
- Give it six weeks. If nothing traces back to posting by then, change the network or the format before you change the volume.
Where launch directories fit
Building in public gathers people who care before your product is ready; a launch gives them a date to act on and gives strangers a way in. Our data has a practical point here: 59.3% of new products link no social account from their own website, so a visitor arriving from a launch or a search result cannot follow the story. Link your account in your site's footer before launch day. If you launch on Awesome Indie, treat the launch as one more update in the series, with what you learned getting there, and run through the launch checklist first.
How we got these numbers
Population. Products on Awesome Indie that passed review (live or accepted for a launch day), submitted between September 16 and November 12, 2026, whose homepage our enrichment pass read successfully: 1,418 products. Snapshot taken November 13, 2026.
How the field is collected. Since August 2026 the submission form has not asked for social links, and since mid-September it asks only for the website. An automated pass then reads the product's homepage HTML and records a social account only if that exact URL appears among the page's links. Makers can add or change links afterwards, and we cannot separate the two sources. A share here means "the product's own site (or, later, its maker) points to this account."
Why only this window. Before August 2026 the form asked makers for social links directly, and the Threads, Bluesky and YouTube fields only arrived in August and September. Comparing years would measure the form change, not makers, so we do not.
Caveats. We read only the homepage HTML, so links rendered by JavaScript or placed on other pages are missed, which undercounts every network. A link says nothing about whether the account is active, or whether it belongs to the product or to its maker. Counting only one product per maker moved no share by more than 0.7 percentage points. Baremetrics counts come from four Internet Archive snapshots and the live page, counting each company listed with a monthly revenue figure.
Sources
- Meta, "Second Quarter 2026 Results Conference Call" (July 2026): https://s21.q4cdn.com/399680738/files/doc_financials/2026/q2/META-Q2-2026-Earnings-Call-Transcript.pdf
- Meta, "Third Quarter 2026 Results Conference Call" (October 2026): https://s21.q4cdn.com/399680738/files/content_files/META-Q3-2026-Earnings-Call-Transcript.pdf
- Microsoft, "Microsoft Fiscal Year 2027 First Quarter Earnings Conference Call" (October 2026): https://www.microsoft.com/en-us/investor/events/fy-2027/earnings-fy-2027-q1
- Space Exploration Technologies Corp., "Form S-1 Registration Statement" (May 2026): https://www.sec.gov/Archives/edgar/data/1181412/000162828026036936/spaceexplorationtechnologi.htm
- Reddit, "Reddit Reports Third Quarter 2026 Results" (October 2026): https://www.sec.gov/Archives/edgar/data/1713445/000171344526000114/earningspressreleaseq326.htm
- Bluesky, "Bluesky's 2025 $100M Series B Lays Foundation for Open Social Web" (March 2026): https://bsky.social/about/blog/03-19-2026-series-b
- TechCrunch, "Bluesky's active user base is shrinking as its focus expands beyond the app" (August 2026): https://techcrunch.com/2026/08/11/blueskys-active-user-base-is-shrinking-as-its-focus-expands-beyond-the-app/
- X, "x-algorithm" repository README (GitHub, updated October 2026): https://github.com/xai-org/x-algorithm
- X, "home-mixer/params/param.rs" (GitHub, November 2026): https://github.com/xai-org/x-algorithm/blob/main/home-mixer/params/param.rs
- X, "Example algorithm diff: bidirectional follow boost" (GitHub, 2026): https://github.com/xai-org/x-algorithm/blob/main/docs/BIDIRECTIONAL_BOOST_CHANGE.md
- TechCrunch, "X cuts headlines from link previews as Musk wants users posting directly on the platform" (October 2023): https://techcrunch.com/2023/10/05/x-cuts-headlines-from-link-previews-as-musk-wants-users-posting-directly-on-the-platform/
- Engadget, "X is testing a new way of opening links in posts to improve engagement" (October 2025): https://www.engadget.com/x-is-testing-a-new-way-of-opening-links-in-posts-to-improve-engagement-211210520.html
- Nieman Lab, "X makes overtures to journalists with new feature designed to improve reach for links" (October 2025): https://www.niemanlab.org/2025/10/x-makes-overtures-to-journalists-with-new-feature-designed-to-improve-reach-for-links/
- The Verge, "X appears to be sending fake traffic across the web" (November 2025): https://www.theverge.com/news/816451/x-fake-traffic-preloading-links-bluesky-substack
- Buffer, "Do Posts with Links Affect Content Performance on X?" (October 2025): https://buffer.com/resources/links-on-x/
- Engadget, "Threads is finally embracing links" (May 2025): https://www.engadget.com/social-media/threads-is-finally-embracing-links-150012499.html
- Engadget, "Threads users still barely click links" (July 2025): https://www.engadget.com/social-media/threads-users-still-barely-click-links-170139103.html
- LinkedIn Engineering, "Engineering the next generation of LinkedIn's Feed" (March 2026): https://www.linkedin.com/blog/engineering/feed/engineering-the-next-generation-of-linkedins-feed
- LinkedIn, "Q4 Earnings & Business Highlights" (July 2026): https://news.linkedin.com/2026/q4-earnings-and-business-highlights
- Bluesky, "Algorithmic Choice with Custom Feeds" (July 2023): https://bsky.social/about/blog/7-27-2023-custom-feeds
- Bluesky, "Introducing Profile Links" (October 2026): https://bsky.social/about/blog/10-22-2026-introducing-profile-links
- Buffer, "Best Content Format on Social Platforms in 2026: 45M+ Posts Analyzed" (March 2026): https://buffer.com/resources/data-best-content-format-social-media/
- Buffer, "Replying to Comments Boosts Engagement by 5-42% on These Major Platforms" (November 2025): https://buffer.com/resources/replying-to-comments-boosts-engagement/
- Buffer, "Introducing our Open Salary System: Reflecting on a Decade of Transparent Salaries at Buffer" (January 2024): https://buffer.com/resources/salary-system/
- Buffer, "Open" transparency dashboard (accessed November 2026): https://buffer.com/transparency
- Baremetrics, "Open Startups" (accessed November 2026): https://baremetrics.com/open-startups
- Internet Archive, "Baremetrics Open Startups" snapshot (June 2018): https://web.archive.org/web/20180625051837/https://baremetrics.com/open-startups
- Internet Archive, "Baremetrics Open Startups" snapshot (January 2020): https://web.archive.org/web/20200109142145/https://baremetrics.com/open-startups
- Internet Archive, "Baremetrics Open Startups" snapshot (January 2022): https://web.archive.org/web/20220104045519/https://baremetrics.com/open-startups
- Internet Archive, "Baremetrics Open Startups" snapshot (January 2024): https://web.archive.org/web/20240110135711/https://baremetrics.com/open-startups
- TrustMRR, "TrustMRR" llms.txt site summary (accessed November 2026): https://trustmrr.com/llms.txt
- TrustMRR, "FAQ" (accessed November 2026): https://www.trustmrr.com/faq
- Management Science, Ryan W. Buell and Michael I. Norton, "The Labor Illusion: How Operational Transparency Increases Perceived Value" (September 2011): https://doi.org/10.1287/mnsc.1110.1376
- Cory Zue, "Farewell, Building in Public" (December 2022): https://www.coryzue.com/writing/building-in-private/
- FlatNine Blog, "Losing transparency" (February 2023): https://flatnine.co/blog/public-vs-private
- The Bootstrapped Founder, "When Your Business Gets Cloned" (February 2021): https://thebootstrappedfounder.com/when-your-business-gets-cloned
- Reddit, "Reddit Rules" (accessed November 2026): https://redditinc.com/policies/reddit-rules
- Google, "URL builders: Collect campaign data with custom URLs" (accessed November 2026): https://support.google.com/analytics/answer/10917952

Diogo builds Awesome Indie, a launch directory for indie makers that has listed products since 2020. He writes about what the data from thousands of launches says about building small software businesses.