Buffer's co-founder wrote roughly 150 guest posts in under a year and credited them with the first 100,000 users. Its blog then shipped four posts a week, counted every trial, and by 2018 drew more than 80 percent of the site's traffic from search. Then revenue fell 20 percent over four years, content was, in the company's word, reignited, and in May 2025 annual recurring revenue hit a new high. Here is the whole run, in Buffer's own numbers.
The Content Machine is a column about startups and scaleups that run content as a system: what they make, how often, how it is reused, where it goes, who makes it, how it is measured and what it moves for the business. Buffer is the only company in the column whose record shows all three acts: the machine being built, the machine being neglected, and the machine being started again. It can show that because Buffer has published its own numbers, including the bad ones, since 2013. Every figure here is Buffer's own unless I say otherwise.
The premise
Buffer sells social media scheduling and publishing for small businesses and creators. It began as a side project in November 2010, joined an accelerator in 2011, and has been fully remote with public salaries since 2013, per its own about page. In 2018 it spent US$3.3 million buying out seven of its sixteen Series A investors, "about half of all the cash we had in the bank", per founder Joel Gascoigne's post at the time. Its December 2025 shareholder update, written by its finance lead Jenny Terry, puts annual recurring revenue at US$23.4 million, 69,760 customers, a team of 74 and 2025 revenue of US$22.5 million.
The machine, in one paragraph. For most of its first decade Buffer's acquisition was a blog about social media, written in-house by two people at a time, shipping about four long posts a week, measured post by post on trials and conversions, and distributed by search. The best of it is on the 2018 post about increasing website traffic, which is also where the company said what the blog was worth.
How it started
Act one is guest posting. Leo Widrich, Buffer's co-founder, wrote "approximately 150 articles for other blogs in less than a year", per Contently's 2014 account, which credited them with about 100,000 users. Widrich's own version to Mixergy was 100,000 users in nine months and a line I have never forgotten: "every article I write equals $500." The original 2012 interview where he gave the figures is behind a wall, so I am citing the two accounts that carry them.
Act two is the blog as a counted machine. Courtney Seiter joined as head of content marketing in April 2014. The January 2014 report Widrich published is the earliest count I have seen a startup make in public: 20 posts that month, "59 total trial starts in January from the blog with 5 conversions", broken down by post type, with the admission that "we've definitely been catching up a bit with our tracking". Kevan Lee, who ran marketing from around 2014, told Scripted in 2015 the Social blog published "four times each week on average, once a day Monday through Thursday", at 1,500 to 2,500 words, 70 percent social media tips and 30 percent content marketing, and that "our content team is me and Courtney". His June 2016 report counted 3,348 sign-ups from the Social blog in the first quarter of that year and 840 from the Open blog, the one where the company publishes its finances and salaries. Brian Peters, in 2018: "more than 80 percent of our website traffic comes from people searching for social media marketing topics", about 1.3 million visits a month, from "only two content writers for most of the past seven years".
How the machine works
Four posts a week, from two people. The cadence was fixed and the team was tiny. That combination is the whole method. A post a day Monday to Thursday, long enough to rank, on two topics the audience would search for, written by people who used the product to promote the post.
Every post scored. Lee tracked unique visitors, time on page, shares, email sign-ups and conversions per post, and scored each one with what he called a variation of Moz's one-metric approach. The quarterly reports grouped sign-ups by source channel. The attribution method is not explained in the reports, and the January 2014 one says the tracking was still catching up, which I would rather read than a clean number with no caveat.
Transparency was a content line, not a value statement. The Open blog published revenue, salaries and the investor buyback, and Lee counted its sign-ups separately: 840 in a quarter in 2016. That is a company turning its own finances into acquisition content and then measuring whether it worked.
Several blogs, one team. Lee's 2016 report lists the formats the six-person marketing team was running by then: the Social blog, the Open blog, an Overflow blog for the engineering side, syndication to other sites, a Medium presence and the email list. Each had its own job and its own count. The Social blog was acquisition. The Open blog was trust. Syndication and Medium were reach for pieces that had already proved themselves at home. The reports read like a small publisher's, because by then that is what the team was.
Search was the distribution. Eighty percent of traffic from search by 2018. Buffer did test paid amplification; Peters reported one boosted post that brought more than 125,000 visits at about six cents a click and "has led to thousands of trials and hundreds of customers". A podcast, The Science of Social Media, launched in September 2016 with weekly interviews.
Then it stopped being tended. The record goes quiet after 2018. Buffer's own shareholder updates give the shape of what followed: ARR peaked at US$22.4 million in February 2020 and fell to US$17.9 million in January 2024, which Gascoigne described in May 2025 as having "dropped 20% in ARR over 4 years". Simon Heaton, now director of growth marketing, told ChartMogul in August 2026 that "SEO and content marketing were a huge part of how Buffer grew in the past" and that "we had to reignite content marketing". A company that had built itself on a blog had let the blog lapse.
What it moves for the business
Act one and two, Buffer's own counts: 100,000 users credited to about 150 guest posts; 59 blog trials and five conversions in January 2014; 3,348 sign-ups from the Social blog in the first quarter of 2016; more than 80 percent of traffic from search in 2018; thousands of trials from one boosted post. All self-reported, all from the people running the blog, and the early ones rest on recollection as much as tracking.
Act three, from the shareholder updates. January 2025: "Blog-driven signups increased 9% thanks to refreshed, high-performing content." May 2025: a new all-time-high ARR on 27 May 2025, the first since the 2020 peak. December 2025: Heaton reported more than 60,000 new customers in a year without increasing marketing spend, per the update's summary. ARR US$23.4 million, up from US$20.7 million at the start of the year.
The confounders are in Buffer's own record too, which is why I trust it. Business of Software's August 2026 account of the turnaround says monthly sign-ups "jumped from 65,000 to 85,000 almost immediately" when the homepage button changed to a free sign-up, which is a product decision, not content. The December 2025 update credits activation, conversion and lifecycle work. The January 2025 update credits a record paid-marketing month in the same breath as the blog. Content was reignited as one of several engines, and the company says so.
What to steal
Count the blog from the first month, badly if necessary. Fifty-nine trials and five conversions, with a note that the tracking was catching up, is the most useful content report I have read from a company of that size. Start counting before the number is good.
Two people and a fixed cadence beat a content team and a calendar. Four posts a week, Monday to Thursday, from two writers, for years. The constraint was the plan.
Turn what you already publish about yourself into acquisition. Buffer's finances and salaries were a content line with its own sign-up count. Your pricing page, your roadmap and your mistakes are content if you count them.
Write down what happens when you stop. Buffer's shareholder updates let me show a 20 percent fall and a restart. Most companies' content lapses are invisible because nobody kept the numbers. Keep them.
Reignite with the old posts first. The 9 percent rise in blog-driven sign-ups came from refreshed content, not new content. The library you let lapse is the fastest thing to restart.
