Skip to content
Analytics

What each network reports back, and what it never will

Why your numbers differ between a tool and the native app, which metrics each network actually returns, and the ones that have been retired for good.

The Picifer team9 min read

Branded Picifer card with the title What each network reports back, and what it never will.

Key takeaways

  • No tool can report a number a network does not return. When a metric is missing from a dashboard, the usual reason is that the API does not have it.
  • Google Business Profile has no per-post metrics at all. The measurements for local post views and button clicks were retired in 2023 with no replacement, so only the location is reported.
  • Follower counts are snapshots, not history. Networks report today's number, so growth has to be measured by storing a value each day rather than asked for after the fact.
  • Tool and native-app numbers differ for honest reasons: different definitions, different time zones, different windows and late updates. Pick one source and stay with it.
  • Measure what you can measure everywhere, and handle the gaps by hand rather than pretending they are zeroes. A blank is information; a zero is a lie.
On this page
  1. The four honest reasons your numbers disagree
  2. What the networks give back
  3. The retirements worth knowing about
  4. Why follower growth only starts when you connect
  5. Measuring what the networks will not
  6. Competitor numbers work the same way, and the same limits apply
  7. Build the reporting habit around the stable layer
  8. The honest version of a dashboard

Open two social media dashboards for the same week and you will get two different numbers. Open the native app as well and you will get a third. The usual conclusion is that one of them is broken, which is almost never the case.

The real explanation is less satisfying and more useful: every network decides what it reports, how it defines each measurement, and when it stops updating. A tool can only show you what the network returns, and the networks do not agree with each other or with their own apps. This guide covers what is actually available, what has been retired, and how to build a reporting habit that survives the gaps.

The four honest reasons your numbers disagree

Before blaming a tool, rule these out. In practice they explain nearly every discrepancy.

Different definitions. Impressions and reach are not the same, and networks define both slightly differently from each other. One counts a view when the post enters the viewport; another counts a scroll past. Our impressions and reach entries cover the distinction, and it is the single most common cause of a mismatch.

Different time zones. A post published at 11pm on Tuesday belongs to Tuesday in one time zone and Wednesday in another. Any daily chart is a decision about which time zone the day starts in, and two tools making different decisions will disagree by one day’s worth of everything.

Different windows. Some networks report metrics for a fixed window after publishing, others report lifetime figures that keep rising. Comparing a lifetime number with a seven-day number makes old posts look better than new ones forever.

Late updates. Networks revise figures after the fact, sometimes days later. A tool that synced on Monday shows Monday’s version of the truth. This is why judging a post on the evening it published tells you how fast it started and not much else.

None of those is fixable, and all of them are manageable with one rule: pick one source for reporting and stay with it. Trends from a single consistent source are worth far more than exact numbers reconciled across three.

What the networks give back

Rather than a false-precision table of every metric, it is more useful to know the shape of what is available and where the real holes are.

Most networks report channel-level numbers well. Followers, impressions or views, engagement counts, and posts published are broadly available across the networks that have a business API at all. This is the layer worth building your reporting on, because it is the layer that exists nearly everywhere.

Post-level depth varies sharply. Some networks return a rich set per post; some return a handful of counts; some return nothing for certain post types. Video is frequently measured differently from a feed post, with its own set of view definitions.

Audience breakdowns are usually a separate permission. Follower counts by country or city, for instance, are often behind an insights permission that a tool has to request and have approved separately from publishing. If a tool does not show them, that is often the reason.

Blogging destinations are the surprise. They are not usually thought of as measurable channels, but a developer-focused publishing platform may report views, reactions and comment counts for your own articles, which is more than some social networks give you.

LayerHow widely availableWhat to do about gaps
Channel totals: followers, impressions, engagements, postsBroadly available wherever a business API existsBuild your regular reporting on this layer
Per-post metricsVaries by network and by post typeCompare only posts from the same network and format
Audience breakdowns and demographicsOften behind a separate permissionTreat as a bonus, never as the basis of a decision
Link clicksInconsistent across networksMeasure at your own site with tracking parameters
Comments and conversationsMissing entirely on some networksReply in the app where no API exists
The three layers of social reporting, how reliable each one is across networks, and what to do when a layer is missing for a channel.

The retirements worth knowing about

Some metrics are not missing from your dashboard. They have been removed from the world, and no tool will ever bring them back.

Google Business Profile post metrics are gone. The measurements for how many times a local post was seen and how often its button was clicked were discontinued in February 2023 with no replacement. What remains is performance data for the location: how people found the listing and what they did next. So there is no way, for anyone, to know how a specific Google Business post performed. That is worth knowing before you plan a reporting routine around those posts, and it is a good argument for putting a tracked link on the button so your own website can tell you.

Google’s business questions and answers API is gone too. It was discontinued in November 2025, again with no replacement, which is why questions on a Google listing cannot be gathered into an inbox alongside the reviews.

Several networks have retired formats along with their metrics. When a format goes, its measurements go with it, and any historical chart that included it becomes a chart with a cliff in it.

Why follower growth only starts when you connect

This one causes more support questions than anything else in analytics, and the explanation is simple.

Networks report your follower count now. They do not generally offer a history of it. So growth is not a number you can ask for; it is a number you compute by storing today’s count and comparing it with a previous day’s.

Which means any tool can only show you growth from the day you connected the channel, and no amount of effort will backfill it. The same is true of a competitor’s follower count: you can snapshot a public account daily and watch it move, but you cannot ask what it was last March.

The practical advice is dull and valuable: connect your channels before you need the reporting, not when a client asks for a quarterly review. A month of stored snapshots is worth more than a sophisticated dashboard with nothing in it.

Measuring what the networks will not

Three gaps come up for everyone, and all three have a workaround that does not depend on a network.

Clicks. Put tracking parameters on every link you publish and measure arrivals at your own site. This is better than a network’s click count in three ways: it is consistent across networks, it survives a network changing what it reports, and it connects a click to what the person did next. Our free UTM builder builds the links, and the UTM parameters entry explains the fields.

Conversions. No network will tell you that a post led to a booking. Your own site, booking system or till can, if the link carries a campaign label. This is also the only honest route to anything resembling social media return on investment.

What happened off-platform. Someone who saw your post and walked in is invisible to every API ever written. The crude and effective answer is to ask. “How did you hear about us” at the counter, on a booking form, in an order confirmation. It is unfashionable and it is real data, which puts it ahead of several things on a dashboard. Our dark social entry covers the broader version of this problem.

Competitor numbers work the same way, and the same limits apply

The instinct, when you want to know whether you are keeping up, is to look up a competitor’s history. It does not exist, for the same reason yours does not: a network reports a public account’s follower count today, not a chart of it.

So competitor tracking is snapshotting. You record a public account’s follower count and how much it posted, once a day, and growth becomes the difference between two days you happened to record. That has three consequences worth knowing before you set it up.

It only starts when you start. No tool can tell you what a competitor’s following was last spring unless something was recording it last spring.

It is limited to what is public. Follower counts and posting frequency, essentially. Nobody outside an account can see its impressions, its engagement rate or its best times, and any product claiming otherwise is estimating or scraping.

Posting frequency is the more useful half. “They post nine times a week and we post three” is a fact you can act on this month. “They have 4,000 more followers” usually is not.

Picifer takes a daily snapshot of a public competitor account’s followers and its posting count for the previous week, which is exactly as much as a network honestly offers. Our share of voice entry covers the broader version of the question and its limits.

Build the reporting habit around the stable layer

Given all of the above, the reporting routine that works is narrower than people expect and more useful.

  1. Pick five numbers, not fifteen

    Per channel: followers, impressions or views, engagements, engagement rate, and link clicks measured at your site. Those five exist nearly everywhere and answer nearly every question you will actually ask.

  2. Use one source and one time zone

    Decide which tool is the record. Note the time zone it reports in. Stop reconciling.

  3. Compare like with like

    Same network, same format, similar kind of post. Comparing a Reel with a link post tells you about formats, not about anything you did.

  4. Respect sample size

    Do not conclude anything from four posts. A finding drawn from a handful of posts should be labelled with how many posts it came from, so it can be read as the weak evidence it is.

  5. Wait a week before judging

    Especially for video. Figures keep moving, and most posts collect a meaningful share of their views after the first day.

  6. Export the raw numbers occasionally

    A CSV of your own data is insurance against a tool’s chart changing, a channel being disconnected, or a network retiring a metric. Picifer’s analytics export is there for exactly this.

The honest version of a dashboard

A good analytics view does three things, and none of them is having the most metrics.

It shows what the networks returned and nothing more, with gaps visible as gaps. It puts the sample size next to any claim it makes about what works, so a conclusion drawn from four posts looks different from one drawn from forty. And it learns from your own results rather than telling you what works on social media in general, because your audience is not an average.

That last point is the whole argument for measuring at all. Generic advice about formats and timings describes somebody else’s followers. Your own numbers, read patiently over a few months, describe yours. Our guide to finding the best time to post is the clearest example of the method, and the social media metrics that actually matter covers which of those five numbers to care about for which goal.

Questions people ask

Why don't my social media analytics match the numbers in the app?
Usually for one of four reasons: the two are counting slightly different things, they are using different time zones to decide which day a post belongs to, one covers a fixed window after publishing while the other counts forever, or the network updated its figures after the tool last synced. None of these is a bug. Pick one source for reporting and stick to it.
Can I see how many people clicked a Google Business Profile post?
No. The metrics for local post views and button clicks were discontinued in February 2023 with no replacement, so no tool can report them. What remains is performance data for the location as a whole: how people found you and what they did. Post-level numbers for Google Business Profile do not exist for anyone.
Why does my analytics tool show follower growth from the day I connected?
Because networks report your follower count today, not a history of it. Growth is calculated by storing the number each day and comparing two days, which means a tool can only show growth from the point you connected. There is no way to backfill it.
Which networks report link clicks?
It varies, and where a network does not report them, the reliable alternative is to measure clicks at the destination with tracking parameters on the link. That has the advantage of being consistent across every network and of surviving a network changing what it reports.
Do all networks provide post-level analytics?
Most do for their own content, but the depth differs sharply, and some provide nothing. Where post-level data is missing, you can still measure the channel as a whole and measure clicks at your own website, which covers most of what a decision actually needs.
How long should I wait before judging a post's numbers?
At least a few days, and a week for video. Networks update figures after the fact, and most posts collect a meaningful share of their views after the first 24 hours. Judging a post the same evening measures how fast it started, not how well it did.
All guides

Put this into practice with Picifer

Picifer plans, writes in your brand voice, schedules to every network and shows you what works. Start free and plan your first week today.

  1. Connect your channels
  2. Teach it your voice
  3. Approve your week

14-day free trial · No card required · Cancel anytime

Start free, no card needed