Content ROI analysis

What a content ROI analysis is

A content ROI analysis is a one-off calculation, usually covering a year, that sets what the content cost against what can be shown to have come back from it. Both sides are built from records you already hold: invoices and timesheets on one side, inquiries and closed business on the other.

The part that makes it worth paying for is the third column. Some of the return can be traced to a page. Most of it cannot be traced to anything, and saying so plainly is the difference between a document you can take to a board and one you cannot. I do not fill that column with an estimate.

It is worth doing when a renewal or a budget decision is coming, when somebody senior has asked whether the content is paying for itself, or before you decide to spend more. It is not worth doing in a first year: there is too little history, most of what will eventually be traced has not happened yet, and you would be paying for a document that says so. Until then, the monthly content report is the right instrument.

What goes into the sum

The cost side is the easier half, and it is almost always larger than the business expects, because most of it is time nobody invoices for.

• Money paid out. Writers, editors, designers, photographers, subscriptions, stock images, the platform itself. Anything with a receipt behind it.
• Time spent inside the business. Hours from the owner, the sales manager, the person who uploads and the person who approves. Counted at a rate you set, not one I invent.
• Maintenance. The updating, the fixing, the answering of comments. Small monthly figures that add up to a surprising annual one.
• What was not done instead. Noted, never costed. If two days a month went to writing rather than to calling lapsed customers, that belongs in the discussion even though it does not belong in the arithmetic.

The return side is built from records of actual business, not from goals configured in an analytics account. Every piece of closed business in the period is examined for whether anything you published touched it: a form on a page, a download, a first email that named an article, a call logged as coming from the guide. Where a record exists, it counts. Where none exists, it does not.

What cannot be attributed

Most of it. That is the honest answer, and it is the reason this document is short and useful rather than long and impressive.

A buyer who reads three of your pages over four months, mentions none of them, and rings the number on an old invoice is invisible. A referral made because somebody once found a guide useful is invisible. And a great many people now get their answer without ever arriving on your site at all, which leaves no record anywhere.

So the analysis reports three figures and never blends them. What the content cost. What can be traced to it with a record behind it. And what closed in the period with no traceable source, stated as a plain number so you can see how much of your business sits outside anything anyone can measure. That third figure is usually the largest of the three, and it is the one most reports of this kind quietly bury.

What you get

The statement. Two pages. Cost, traced return, untraced business, and the method used for each, written so a sceptical accountant could follow it.

The working file. The spreadsheet behind the statement, with every piece of closed business in the period and an evidence column either filled in or deliberately left blank. You keep it, and next year's analysis is half the work.

The judgment. One page saying what I think the numbers support: continue, cut, change shape, or stop. Where the evidence does not support a judgment, it says so, and says what would have to be recorded from now on to make next year's answer clearer.

A worked example

An illustration. A bookkeeping firm in Halifax has published for three years: forty-one articles, a monthly newsletter and six service pages. The cost side turns out to be roughly twice what the firm believed, because only the freelance invoices had ever been added up and the managing partner's own writing time had never been counted at all.

Of thirty-one new clients in the year, nine can be traced: seven arrived through a form on two of the service pages, and two named an article in their first email, which the office manager happened to have recorded. Twenty-two have no traceable source. The statement says exactly that, with both numbers stated plainly. The judgment is that the two earning service pages deserve the budget the forty-one articles have been absorbing, and that from January the intake form should ask every new client where they heard of the firm. That single question will do more for next year's analysis than any amount of tooling.

How it runs

1. A half-hour call. What decision the analysis is meant to inform, and what records exist. If the records are too thin to support an answer I say so on the call and we stop there. No charge.
2. Records. Invoices for the period, a rough hours estimate from each person involved, your client list, and read-only access to the inquiry log.
3. The reconciliation. Five to eight working days. Most of it is going through closed business one line at a time, which is slow and cannot be automated.
4. The walkthrough. Forty-five minutes on video with whoever holds the budget. The statement arrives the day before, so it is read cold first.

What it costs

There is no published band for this one, because the work scales with how much closed business there is to go through and how well it was recorded. It is quoted in CAD as a fixed price after the half-hour call and fixed in writing before anything starts. The pricing page shows the four standing ways of working; this sits alongside them rather than inside one.

What happens next

The most common outcome is not a spending decision but a recording decision. Businesses discover they cannot answer the question and start capturing the one or two fields that would let them, which means agreeing a small set of measures and running a monthly report against them. Where the analysis shows part of the archive earning nothing, the follow-on is clearing out the pages that are not. None of that is assumed. The analysis is complete on its own, and its job is to answer one question honestly.