Content ROI Calculator

What the formula says

Content return is the contribution attributable to content divided by its cost: (attributed contribution − production cost) ÷ production cost.

The complication is that content costs are incurred once and returns accrue for years, so a single-period calculation understates it badly.

The cumulative version, total contribution to date against total production cost; is the more honest measure for anything with a long shelf life.

The numbers, worked through

An article costing $600 to produce and promote, generating 340 organic sessions a month at a 1.8% conversion rate: 6.1 orders monthly, $195 of contribution.

First-month ROI is −68%. By month four it has repaid the cost. Over three years, assuming traffic holds, it generates roughly $7,000 of contribution, an ROI of 1,067%.

Judged in the month it was published, content always looks like a poor investment. Judged over its life it frequently outperforms every paid channel.

The difficulty is that the same arithmetic applies to content that never gains traction, and the failures cost the same to produce as the successes.

What the number leaves out

Attribution to content is weak, since a reader may return through search or direct traffic weeks later and be recorded elsewhere. Last-click measurement of content is close to useless.

Traffic also decays unless maintained, so a three-year projection built on current traffic assumes updates that may not happen.

Turning it into a decision

Measure cumulatively at the page level, tracking each piece's traffic and assisted conversions over its life rather than reporting content as a monthly channel.

Then judge the programme rather than the piece. Content follows a power distribution, a small share produces most of the traffic, so an average across all pieces describes neither the winners nor the failures.

Why the distribution matters more than the average

In most content programmes, roughly a fifth of pieces generate the large majority of organic traffic, and a substantial share generate almost none.

That means the programme's return depends on producing enough pieces to find the winners, and on identifying and updating them once found. Refreshing a piece that already ranks is consistently higher-return than producing a new one.

It also means judging content by average performance leads to abandoning the activity, since the average is dragged down by the majority that fail. The correct comparison is total programme contribution against total programme cost, which for a well-run content operation is usually favourable even though most individual pieces are not.

Updating existing content generally outperforms creating new content on a return-per-hour basis, since a page already ranking has demonstrated demand and needs less to improve than a new page needs to start.

Auditing the existing library for pages that once performed and have declined is usually the highest-return content activity available, and it is the one least often scheduled.

Internal linking from established pages to new ones materially shortens the time a new piece takes to rank, which improves the return on every subsequent piece produced.

Assisted conversions are the honest way to credit content, since last-click measurement will attribute almost none of its value and will lead to the programme being cut.