Gross Retention Rate: The Number That Cannot Be Rescued by Upsell

Gross retention rate is the share of your opening recurring revenue that is still with you at the end of the period, after cancellations and downgrades and before any expansion. It answers one question and refuses to answer any other: of the money you already had, how much did you keep.

That narrowness is the value. Almost every other growth metric can be improved by selling more. Gross retention cannot. If customers are leaving or shrinking, the number falls, and no amount of upsell to the survivors will lift it. It is the closest thing a subscription business has to a plain statement about whether the product delivered what the sale promised.

The calculation

Gross retention equals opening recurring revenue minus churned recurring revenue minus contracted recurring revenue, divided by opening recurring revenue, times one hundred. Opening recurring revenue is the total for the customers on your books on day one, frozen at that moment. Churned recurring revenue is the amount lost from accounts that ended entirely. Contracted recurring revenue is the size of the reductions among accounts that survived at a lower value. Expansion appears nowhere.

A worked example

A company opens the quarter with 20,00,000 in monthly recurring revenue. Over the quarter, accounts worth 60,000 cancel outright. A further group of accounts renew at lower values, reducing recurring revenue by 20,000 in total. Separately, surviving accounts expand by 1,10,000, and new customers bring in 3,40,000.

The numerator is 20,00,000 minus 60,000 minus 20,000, which is 19,20,000. Gross retention is 19,20,000 divided by 20,00,000, or 96 per cent. Note what happens if the two excluded items creep in. Adding expansion gives 101.5 per cent, which is net retention. Adding new business as well gives 118.5 per cent, which is not a retention metric at all, merely growth wearing a retention label.

Reading down the first column explains most of the confusion in retention reporting. Three of these four metrics answer questions about the existing base, but only two of them are capped, and only one of the capped pair is measured in money. That one is gross retention, and it is why it tends to be the first number an experienced reader looks for.

The most effective way to inflate gross retention without stating anything false is to measure only the accounts that were up for renewal. On annual contracts this seems reasonable, because accounts that were mid-term never had a chance to leave. The problem arises when the resulting figure is presented as if it described the whole book. A quarter in which few large contracts came up will produce a strong number that reflects the calendar rather than customer behaviour.

The second method is excluding contraction. A customer who renewed at half their previous value is counted as fully retained because they are still a customer. This can run for years. The base steadily shrinks at every renewal, gross retention keeps reporting a comfortable figure, and the problem only surfaces when the accounts eventually leave from a much smaller starting point.

The third is a moving denominator. Recalculating opening recurring revenue from a live table means that any backdated correction, late-recorded cancellation or currency adjustment quietly rewrites history. If a retention figure from four quarters ago cannot be reproduced exactly today, the series is not measuring change over time, it is measuring the state of the database on the day the query ran.

Good gross retention is steady rather than spectacular. It holds within a narrow band across quarters, it looks similar when segmented by plan and by industry, and the reasons recorded against the losses are the ordinary ones: the business closed, the sponsor left and the successor had different priorities, the need genuinely changed. Nothing about it depends on a single large account behaving well.

Bad gross retention has recognisable signatures. A drop concentrated in the two quarters after a price increase suggests the increase outran the perceived value. Losses concentrated in one acquisition channel suggest that channel delivers customers who were never a fit, which is a sales and marketing question rather than a product one. And a pattern of contraction without cancellation is the quietest failure of all, because the metric declines slowly enough that no single quarter ever looks alarming.

Gross revenue churn is the same measurement inverted, so the two must always agree to one hundred per cent. Net revenue retention adds expansion and therefore has no ceiling. Logo retention counts customers instead of revenue and will disagree with gross retention whenever account sizes vary. Contraction is the downgrade term inside the calculation, and expansion is the term deliberately kept out of it. When someone quotes a retention figure without saying which of these they mean, the safest assumption is that it is the most flattering one available.

Continue with the related metrics and the tooling that keeps them reproducible:what a CRM is, AI CRM, lead management software, sales automation, CRM for small business, and features overview.