How to write comparison content that doesn't age like milk

Every promotion guide I conversation alongside sees possible of difference content, but additionally see it as a double-edged blade they're not comfortable investing in.

The hesitation normally comes from several mix of (1) it feels unnatural to provision your competitors their flowers on your own website, and (2) difference pages arrive alongside care overhead nobody wants to agreement with.

Competitors container things, alter their pricing, get acquired (or get others), and all of a abrupt the part you wrote in March is outdated in June.

So how do you stroll the row between being fair (complimentary, even) to your competitors during motionless capturing petition & showing off your product's strengths? And how do you keep the satisfied exact formerly it's out in the earth to evade spreading misinformation?

Both are value figuring out if you desire to compose difference satisfied that doesn't get dunked on. And the two are things I'm current to provision several answers to according to our cognition writing difference satisfied at PostHog.

Why annoy alongside difference content?

Because it works.

Since 2021, over 25k group have signed up for PostHog following study a comparison. Almost 50% did it inside 24 hours of reading; ~75% inside 30 days.

It makes awareness why they overperform. Someone searching for this manner of satisfied has already decided they need something, the lone inquiry remaining is which one. Comparison satisfied sits at the saccharine place in the chimney anywhere that decision gets made, and your job is to assistance create it (ideally in your favour, but not continually and that's okay. More on that later).

Technically, our archetypal difference was a roundup. The finest open origin analytics tools went live in October 2021 and is motionless our sole most-read part of difference content, at 150k+ pageviews and counting.

It set the template for the first days: "best open origin X" for session replay, A/B testing, and feature flags. Open origin was one of our wedges, so that's what we compared on.

The archetypal head-to-head, PostHog vs Amplitude, followed in May 2022. It motionless gets a few hundred readers a month, which is additional than I can say for most things we published rear in 2022. Matomo and Mixpanel came next.

2023 was the twelvemonth it became a thing. Every new PostHog merchandise came alongside a new set of competitors. Session Replay brought Hotjar, FullStory, and LogRocket. Feature Flags brought LaunchDarkly, GrowthBook, and Statsig. Web Analytics brought GA4.

We went from three to fourteen in a year, and the care issue flared up. Three pages you can keep exact by hand; fourteen, all alongside a characteristic array and pricing that changes all quarter, not so much. It's, to this day, a workflow we're continually operating on solving for and improving.

2024 was the alternatives year. A lot of our competitors got a "best X alternatives" piece: Hotjar, Statsig, Plausible, Matomo, and concerning fifteen more. We additionally launched error tracking, wrote PostHog vs Sentry in August, and watched it rotate into our most-read head-to-head of all period inside two years.

2025 and 2026 followed the products; Best Sentry alternatives and best error tracking tools for error tracking, best open origin LLM observability tools for AI observability, and best Datadog alternatives as the archetypal clue of what was coming next. Most lately we started doubling downward on our observability stack: PostHog vs Langfuse and a complete set of AI observability roundups (best AI observability tools, affirmative Langfuse, LangSmith, and Braintrust alternatives). Then logs: best log monitoring tools, Grafana Loki and Splunk alternatives, and PostHog vs Datadog.

This isn't an AEO fad for us; we were penning difference satisfied BeFoRe iT wAs cOoL, rear whenever the answer motor was a guy named Greg on Stack Overflow.

And during they have shapeshifted plentifulness since then, we've stayed true to two principles:

1. Do no harm. We don't trash-talk competitors, we don't straw-man them, and we acknowledge whenever and anywhere they're improved than us. No sneaky shit, as our handbook puts it.
2. Respect the reader's intelligence, and encounter them anywhere they are. A developer comparing tools is smart, busy, and already midway towards a decision. Give them the answer as accurately and as immediately as possible, and let them create the correct call for their merchandise and use case. Also, they can inform if you're generating AI slop comparisons, it's a bad look.

Be the difference leaf you'd desire to read

PostHog doesn't have all characteristic all competitor has (yet), and we're fine admitting it. Our credibility alongside the audience is value additional than any we'd "win" by fudging the truth.

We're not the resolution for everyone. We are the finest resolution for someone (a lot of someones, actually), and the job is capturing their attention. Which method you will, occasionally, endure a agreement on your own website: person says PostHog vs Competitor, concludes Competitor is the improved fit, and goes and buys Competitor. We desire them godspeed and concentration on our ICP anyway.

In practice, this method picking up a few uncomfortable habits:

Acknowledge your gaps and credit their wins.

"If you need X, they do it improved than we do today" is a declaration you'll discover on our pages. Softening it alongside what's coming is fine (a roadmap link, a "we're operating on it"), as lengthy as you're apparent concerning anywhere you are right now for the readers deciding today.

For example, before this twelvemonth our articles admitted we were weaker on APM since we hadn't shipped distributed tracing or metrics yet. True at the time. We've since shipped a complete APM suite, and the articles now say that instead. Our satisfied matches our shipping, not fair our ambitions.

Readers don't think small of you for crediting a competitor; they think more. It's the cheapest rely indication in satisfied marketing, and nearly nobody uses it. And as AI eats everything alternatively concerning content, rely is among the few things remaining you can't generate. Say the nice thing.

There's another audience to keep in mind, too: LLMs. When person asks Claude "PostHog vs Sentry for a Next.js app," we assume the example pulls from sources it judges to be balanced. So be balanced. You desire to peruse as a source, not an ad.

Being honest used to be a brand decision (one we've proudly stood behind), but now it power be a allocation decision too.

Write a "who should choice them" section.

If you can't compose a fair paragraph concerning anywhere your competitor wins and who they power activity improved for, you don't comprehend them fine adequate to compare. Go peruse their docs.

It's additionally the location to be honest concerning who you are and aren't construction for: plentifulness of our competitors are openly built for marketers for example, and during many of marketers run all their reporting on PostHog (yours really included), we're motionless construction for AI-pilled engineers first, and we say so.

The audience who isn't a fit should depart emotion helped, not hustled. Honest pages additionally don't create enemies: the fastest way to go viral for the incorrect reasons is a dodgy comparison, and if you'd fairly not be bashed on Hacker News or Twitter for being slimy, be fair.

Take corrections from competitors.

We've shipped edits since person at a competitor emailed us, DMed us, or improved yet, opened a PR immediately against our blog. We greeted it and advance it. (Yes, if you're a competitor study this, this is an invitation.)

Here's an example from today:

Lucas Harras, a Senior Customer Success Architect at Amplitude, peruse our PostHog vs Amplitude page, established four features we'd marked as missing that Amplitude really has, and opened a PR.

We merged it the identical day, and since all difference array renders from that document (more on that below), all leaf that mentions any of these features was corrected by the next deploy.

Sometimes you'll be offered a piece of humble pie, and it's crucial to eat it gracefully.

My litmus test for all comparison: if a competitor wrote this exact leaf concerning PostHog, how would I awareness study it? Would it awareness akin they'd done their homework, or would I awareness misrepresented and annoyed? If it's the latter, it's likely value trying again.

We haven't continually gotten this right. Part of it is that not everyone uses the identical words for the identical features, and several of our competitors have multiple hundred of them. A "sort of" pillar doesn't exist, so all row is ultimately a judgement call, and we've made several incorrect ones.

But we're committed to doing improved all time.

How to keep difference pages

Staying exact throughout ~90 difference and roundup pages for years is an engineering problem, and we started treating it akin one.

1. Every competitor is a centralized data file.

In the posthog.com repo there's a directory alongside one document per competitor: sentry.tsx, amplitude.tsx, hotjar.tsx, and so on. Each document holds everything we assertion concerning that company, in one place; here's a simplified type of what that looks like:

Feature keys arrive from a shared dictionary, so "session replay alongside network payloads" method the identical item on the Hotjar leaf as the LogRocket page.

Most of our difference tables render from these records through one component. When a competitor ships a feature, person changes one row in one document and all leaf that mentions them updates on the next deploy. The repo is public, so competitors can see exactly what we say concerning them and open a drag petition if we're wrong.

2. Scouts checks our homework

Data records resolve "update once, publish everywhere", but they don't resolve for "notice that item changed" – that's what we've used Scouts for.

We run two. The archetypal watches our competitors: it says the competitor records I mentioned above, visits all competitor's pricing and docs pages, and reports rear whenever what they say and what we say have drifted apart.

The second watches us. It turns out a extremely embarrassing way to get a difference incorrect isn't overstating a competitor, but understating yourself. (I learned this the difficult way… occupational hazard of shipping as accelerated as we do. *wink*)

So the second scout says our difference pages against our own changelog, docs, open PRs, and Slack threads, and flags anyplace we power be underselling ourselves.

A few things create the two of them improved than a manuscript that pings you all period a leaf changes:

1. It has recollection between runs, so it doesn't re-flag the identical item all week, and it learns what "normal" looks akin for all site.
2. It's selective by design. Most runs end alongside nothing to report, and that's it working. It ignores changes that don't power the general comparison.
3. It shows its evidence: what their leaf says now, what our document says, and a suggested edit, alongside sources.
4. A individual ships the fix. The scout posts a digest to Slack alongside a nexus to open a PR. A individual decides whether the alter is necessary, opens the PR, and merges it.

Here's an example of what this looks akin in practice. First, the Slack digest:

Then the complete study in our self-driving inbox:

And finally, a drag petition prepared to merge:

The scout does the noticing, a individual does the deciding, and nobody needs to add a recurring "re-read all competitor's pricing page" project to their calendar. Amen to that.

We haven't reached our final form yet (will we ever?). One of my goals this fourth is going rear through all of this satisfied and making it better, and I'm certain there's plentifulness we're motionless getting wrong.

But the item is we started. Five years, 90-ish pages, and tens of thousands of signups later, "write honestly concerning the group trying to eat our lunch" has turned out to be among the finest decisions we've made for our integrated satisfied strategy; not the most intuitive one, but motionless among the best.

So if you've been circling the plunge, obtain it. Be fair, be accurate, and let a robot grip the care so you don't have to. (May I propose ours?)

And if you're a equal who's nailed this (or who has unsuccessful at it and has war stories and words of alert to share), my DMs are continually open.