Every multi-platform competitive review runs into the same wall. You have a competitor's numbers on three platforms, and the obvious thing to do — put them side by side and see who's winning where — is the one thing the data won't support.
The wall is real. The way around it isn't to give up on cross-platform analysis; it's to compare different things.
Why the numbers don't compare
Covered properly in why cross-platform averages lie, but the short version:
Platforms don't agree on what a view is. They differ on the watch threshold, on whether replays count, on impressions versus plays, on how autoplay is treated. They don't agree on what an engagement is either — which reactions exist, whether saves are exposed, whether private shares are counted.
So a 6% rate on one platform and a 2% rate on another are not two measurements of the same quantity. They're two measurements of different quantities that happen to share a name.
Layer on top of that: discovery-led feeds show content mostly to strangers, subscription-led surfaces mostly to followers, and strangers convert to engagement at a much lower rate. A platform can deliver more reach, more new audience and a worse rate, all from identical content.
What does compare: five reads
1 · Effort allocation
Count pieces per platform per month, for each competitor. This is a straight count, so it's directly comparable — no rates involved.
You're reading their bet. A competitor publishing 20 short pieces a month and one long piece a quarter has made a decision about where their audience is, and it's a decision informed by data you can't see. Their allocation is a compressed summary of their analytics.
Watch the trend more than the level. Allocation shifting towards a platform means something is working there.
2 · Format adaptation
The most revealing read available, and the cheapest.
For each competitor, take one piece of content that appears on multiple platforms and compare the versions. Three possibilities:
Identical upload everywhere. They're distributing, not making platform-native content. Usually a resourcing decision and usually underperforming on at least two of the three.
Recut per platform. One shoot, several edits — different lengths, different openings, different aspect ratios. This is the common professional pattern.
Genuinely different content per platform. Separate concepts, separate shoots. Expensive, and a sign of either real conviction or an uncoordinated team. You can usually tell which from whether the pieces share a message.
Where a competitor sits on this scale tells you their production maturity more reliably than production value does.
3 · Topic travel
Take a topic and follow it across platforms.
Does their pricing content exist everywhere, or only on YouTube? Is the customer-story format short-form on one platform and long-form on another?
Disagreement is the finding. A topic that performs on one platform and not another — for them or for you — is telling you something about format fit and audience intent that neither platform's numbers alone would surface. That's a genuine cross-platform insight, and it doesn't require a shared unit.
4 · Audience overlap
Ask, from public signals, whether it's the same audience three times or three audiences.
Look at who comments, whether the same handles recur across platforms, whether the tone of the comments differs, whether the questions being asked are at the same stage of the buying process.
This changes what the numbers mean. Three platforms reaching the same 20,000 people is a very different business from three platforms reaching 60,000 distinct people, and the headline totals look identical.
5 · Same-platform head-to-head
The one comparison that's numerically valid: your YouTube against their YouTube, over the same window, on the same metric.
Do it three times, once per platform, and present three separate results. Never total them. This is the backbone of share of voice and it's the only cross-competitor number in this article that means what it appears to mean.
Presenting it without lying
The format that works is one section per platform, then a synthesis section that contains no numbers.
The synthesis is the deliverable and it survives translation. Notice it also names what the analysis didn't settle, which is the part that usually gets dropped and is usually the most important line on the page.
Three traps
The blended scorecard. Any table with a "total" or "average" row across platforms. It measures your posting mix, not your performance.
Treating absence as weakness. A competitor with no TikTok presence may have tested it and found their buyers aren't there. Absence is a decision, and it might be the right one — the same four reasons apply as for any other gap.
Judging platforms by their own house metric. Every platform's dashboard is built to make that platform look essential. Judge each against what you actually need from it, which requires having set a goal first.
Where Acumin fits
Competitor reads are per platform by construction, and comparisons are same-platform against each competitor's own distribution. There's no blended cross-platform score in the product, deliberately — it would be the exact category error this article is about.
Cross-platform work today is strongest on YouTube, which has the richest public data. Instagram and TikTok reads are shallower because the public surface is thinner, and they're labelled as such rather than padded out to look symmetrical.
How to use this today
Pick your closest competitor and one piece of their content that exists on two platforms. Watch both versions.
If they're the same file, you've learned they're distributing rather than adapting — and that's an opening. If they're genuinely recut, look at what they changed in the first three seconds. That edit is the most concentrated piece of platform craft you can observe for free.
Related: Why cross-platform averages lie is the underlying reason the numbers won't blend. Share of voice without a media budget is the per-platform head-to-head done properly.