A link your client opens whenever they wonder

The monthly report is the wrong artefact

Almost every agency produces a monthly report, and almost every agency privately knows it is not doing the job. It takes hours to assemble, it lands in an inbox, it is skimmed or not opened, and by the time anybody discusses it the data is three weeks old. Meanwhile the client — who has been wondering how things are going since the middle of the month — has had no way to find out.

The format is a relic of when the data was genuinely hard to reach. When pulling together rankings, traffic, spend and leads meant exports from four systems and an afternoon in a spreadsheet, doing it once a month was the only sustainable cadence. That constraint has been gone for a decade and the ritual has outlived it.

The cost of the ritual is not only the hours. It is that between reports, the client has no visibility, and a client with no visibility fills the gap with anxiety. The characteristic pattern is a quiet three weeks followed by an email that begins "we were wondering whether" and ends in a conversation about value. That email is nearly always caused by the reporting cadence rather than the campaign performance, and it arrives at the moment the agency has the least fresh data to answer it with.

Worse, the monthly report is where an agency's worst month gets its full dramatic weight. A bad fortnight buried in a live view that the client checks casually is a bad fortnight. The same fortnight presented in a formal document, once, with a chart, is an event. Agencies who move clients to live reporting consistently report that conversations about dips become smaller and more practical, because the client saw it happening and saw it recover.

It also concentrates all the client's attention on one artefact you spent hours making and they spend four minutes reading. The hours are real: across twenty accounts, a monthly report that takes ninety minutes to assemble is most of a working week, every month, spent producing documents rather than results. That is the single largest block of non-billable time in most agencies and it buys remarkably little.

The replacement is not a fancier report. It is a link. The client opens it when they wonder, sees this month to date in their own timezone with last month beside it, and gets an answer without asking anyone. The scheduled email still goes out, because people like a prompt, but it is a summary that points at the live view rather than a document that has to carry everything.

What the client sees

The layout is chosen for somebody who runs a business rather than somebody who runs campaigns. Everything below the first screen is optional detail.

How reporting gets set up and then stops needing you

The intended end state is that nobody on your team touches reporting in a normal month.

1. 1

   Choose what this client sees

   Sections are switched on per client. A client who only buys SEO does not need an ad spend panel, and a client who has never asked about cost per lead may be better served without one.

2. 2

   Brand it once

   Logo, colours, domain and sending address are set at the agency level and inherited by every client, with per-client overrides where a sub-brand needs them.

3. 3

   Share the link

   A single URL, optionally password protected, that always shows current data. No login to create, no credentials for the client to lose, no support ticket about a forgotten password.

4. 4

   Set the schedule

   Weekly, monthly or quarterly, to whichever recipients you name, from your address. Sent on the client's calendar so a monthly report covers their month rather than the server's.

5. 5

   Stop doing it

   That is the whole point. If somebody on your team is still assembling a report by hand in month three, something has been configured wrong.

Reporting as the retention mechanism it actually is

Ask a room of agency owners why clients churn and you will hear about results, about price, about the client's budget cycle. Ask them to describe the last three accounts they lost and the pattern is usually different: the client went quiet, then asked for a call, then said they were "taking things in-house for a while". Almost none of them produced a specific complaint about performance, because the client did not have enough visibility to form one.

That is the mechanism worth designing against. A client who can see leads arriving, from known sources, at a known cost, does not spend three weeks quietly wondering. The wondering is where cancellations are born, and it is remarkably cheap to eliminate.

There is a second-order effect that matters more. A client with a live report starts engaging with the data, and an engaged client asks better questions. "Why did calls drop in the second week" is a question you can answer and that leads somewhere. It is a far better conversation than "we are not sure this is working", and the difference between the two is entirely a function of whether the client has been looking at numbers all month or looking at nothing.

It also changes who inside the client's business is looking. A monthly PDF reaches one person, usually the owner, usually while they are busy. A bookmarked link gets opened by the office manager, the sales lead and whoever answers the phone — and those are the people whose behaviour actually determines whether your leads convert. An agency with three people at the client watching the same numbers has three advocates rather than one budget holder.

Agencies sometimes resist this out of a worry that visibility invites scrutiny. It does. The scrutiny arrives either way; the only question is whether it arrives continuously in small pieces you can address, or all at once in a meeting called to discuss the relationship. The first is survivable and the second frequently is not.

The one genuine caution is that live reporting punishes agencies whose numbers do not hold up. If the lead count only looks good because it includes spam and existing customers, giving the client a list of individual leads to click through is going to expose that in week two. This is a feature. It is also a reason to get the underlying measurement right before turning the link on.

Live link, scheduled email, exported document

Three formats, three jobs. Most accounts want the first two and use the third occasionally.

The details that decide whether a report is believed

A client report is a trust artefact, and trust is lost through small inconsistencies long before it is lost through bad results. Several of these are unglamorous enough that most reporting tools get them wrong.

The month has to be the client's month. A report generated on server time will, for any client in a different timezone, include or exclude several hours at each boundary — which means the client's own count and yours disagree by a lead or two, every month, forever. That is a small error that destroys confidence out of all proportion to its size, because the client cannot tell a boundary artefact from a system that loses data.

The chart and the total have to agree. Bucketing a daily chart in UTC while computing the total on the client's calendar produces a report where adding up the bars does not give the headline number. Someone will eventually add up the bars.

The same lead must not appear twice. A visitor who fills a form and then rings is one person; counting them as two inflates every number downstream and is obvious to any client who recognises the name in the lead list. Recognising a name in the list is exactly what clients do first.

The comparison period has to be the right one. Month-on-month looks like the obvious default and is misleading in every seasonal business, which is most local services: a roofer in February against a roofer in January says nothing useful. Year-on-year is the comparison that means something, and it is worth showing both so the client can see the seasonal shape rather than being alarmed by it annually.

And the definitions have to be stated. "Leads: calls over 30 seconds, form submissions, and click-to-call taps, excluding spam and repeat callers within 30 days" is a sentence that prevents an enormous amount of argument. Reports that show a number without saying what it counts invite the client to assume it counts whatever makes the agency look best, and they are often right.

What makes this work across an account base

The difference between reporting for one client and reporting for forty is almost entirely about what happens without anybody doing anything.

• Schedules that actually fire. Weekly, monthly and quarterly all run through one sender, so a quarterly report does not go out once and then silently stop — which is a real failure mode and an embarrassing one.
• Branding inherited, not repeated. Set at the agency level and applied to every client, with overrides available. Nobody configures a logo forty times.
• The client's timezone, everywhere. Totals, charts and schedule boundaries all use it. A report is one of the few places where a timezone bug is visible to a customer.
• Access is scoped server-side. A shared link resolves to exactly one client's data. Changing an identifier in the URL reaches nothing.
• Links can be revoked. A shared report link can be expired or rotated, which matters when a client's staff change or a relationship ends.
• Delivery is verifiable. You can see that a scheduled report was sent and when. "We never got it" deserves a better answer than "it should have gone".

What to do about a bad month

Live reporting means the client sees the bad months as they happen, which is the objection agencies raise most often and the one worth answering directly. The honest answer is that it changes the conversation rather than removing it, and the changed version is better in almost every case.

Under monthly reporting, a bad month arrives as news. The agency has known for three weeks, has been working on it, and now presents both the problem and the fix at once — which reads, from the client's side, exactly like being told late. Even when the agency handled it perfectly, the client learns that something went wrong and that they were the last to know. That is a trust cost paid every time.

Under live reporting, the client may notice the dip before you mention it, which sounds worse and is generally not. What actually happens is that you get there first, because you are looking at the same data more often: a note on the eighth saying calls are running behind and here is what we are doing carries far more credibility than a paragraph on the first of the following month. The reporting cadence forces a communication cadence, and the communication is what clients are actually buying.

The practical discipline is to never let the client be the first to raise a dip. That requires somebody on your side looking at the cross-account view weekly — which of your accounts are behind where they were, and why. Fifteen minutes across forty clients, once a week, prevents most of the conversations agencies dread. And when a dip is genuinely outside your control — a client's phone system down for three days, a seasonal collapse, a competitor opening — having said it first is the difference between an explanation and an excuse.

How it fits with everything else

Reporting is the surface of the whole system. It shows tracked calls, captured forms, attributed journeys, cost per lead, rankings from the SEO suite and reviews from reputation — which is why it can assemble itself. Every number on the page already exists somewhere in the same database, computed the same way, on the same clock.

That is the real argument for one platform over five integrated tools. Not that integrations are hard, but that a report built from five sources has five definitions of a lead, five timezone assumptions and five monthly boundaries, and reconciling them is a job somebody has to do every month forever. Reports assembled from one source are internally consistent by construction.

There is one more consequence worth drawing out. Because the report is generated rather than assembled, it costs the same whether you have four clients or forty. Reporting is normally the cost that scales fastest and worst in an agency — the one that means winning ten more accounts requires another account manager purely to produce documents. Removing it changes what an agency of a given size can carry, which is a more consequential effect than the reporting itself.

The white-label layer wraps all of it. On plans that include it, the client sees your domain in the address bar, your logo on the page, your address in the sending header, and no mention of any vendor anywhere. That is not vanity — it is what makes the reporting look like something your agency built, which is what the client is paying for.

Common questions