What is Backlog?

Backlog is the most reassuring number in a services business: revenue already sold, contracted, waiting to be delivered. Reassurance is exactly why it deserves suspicion. Two firms can report identical backlogs and face entirely different futures, because the headline number carries none of the information that decides what the backlog will actually become.

Contracted future revenue

Backlog is the value of work a firm has contracted but not yet delivered, and therefore not yet recognised as revenue. It sits between two better-understood quantities: pipeline, which is work the firm hopes to win, and revenue, which is work already performed. Pipeline is hope, revenue is history, backlog is obligation: the order book of a people business. The edges of the definition are soft, and the softness matters. Whether a firm counts unexercised contract options, assumed renewals, or the full value of framework agreements under which nothing has yet been ordered can move the number dramatically, and pressure to show a healthy backlog tends to move it in one direction only. Before comparing backlogs, or trusting one, establish what was allowed in.

Quality beats size

A backlog’s size answers one question: how much have we sold? Its quality answers the one that matters: what will this backlog become when delivery meets it? Three dimensions decide that.

• Embedded margin. Backlog sold thin is a queue of low-profit obligations. Work priced aggressively in a competitive year does not become profitable by sitting in the order book; it waits there, and the firm’s future delivery margin is largely written before delivery begins.
• Deliverability. Contracted revenue only converts if the firm can staff the work, at the promised time, with the promised skills. Backlog scheduled into quarters where the practice is already full does not convert on time; it slips, strains clients and invites penalties. This is a question of delivery confidence, engagement by engagement.
• Concentration. Backlog dominated by one client, one programme or one renewal decision is one procurement meeting away from being a different number. Contracts have break clauses; programmes get descoped. The thicker the concentration, the more the backlog is a bet on a small number of other people’s decisions.

Backlog size measures how much has been sold; backlog quality decides how much of it will ever become revenue, at what margin, and on whose timetable.

Burn and replenishment

Backlog is a reservoir: bookings fill it, delivery drains it. The drain is backlog burn, the rate at which contracted work becomes delivered revenue; the fill is replenishment, the rate of new bookings. The two rates matter more than the level. Strong burn without replenishment is a firm eating its order book: revenue looks excellent right up to the cliff. Strong replenishment without burn means selling faster than the firm can deliver: the backlog ages, promised start dates slip, and deliverability decays while the headline grows. The ratio between the two flows is a metric in its own right, the book-to-bill ratio, and a rough sense of runway comes from expressing backlog in months of current revenue. Purely as illustration, a backlog equal to nine months of revenue sounds like nine months of safety, and it is only that if the margin, deliverability and concentration behind it hold.

One concrete example

Clearly illustrative, no customer implied. A firm of a few hundred people reports a backlog equal to about nine months of revenue and treats it as proof of a safe year. Composition tells another story. A third of it is a single multi-year client with a break clause a few months out. Another slice was sold at thin margin during a bruising competitive cycle. A further portion is scheduled into quarters where the relevant practice is already fully committed, so it will slip regardless of appetite. Nothing in the headline distinguishes this backlog from one of identical size made of diversified, well-priced, deliverable work. The first firm holds nine months of assumptions; the second holds nine months of revenue. Both report the same number.

The decision-intelligence reading

Read as a decision input rather than a comfort, backlog should trigger interrogation, not celebration. Each component of the number is a claim, and the claims come at different evidence strengths: a signed contract with committed dates is close to measured; an assumed renewal is stated; the staffing plan that makes a start date credible may be measured, modelled or merely hoped. The decision-intelligence discipline is to attach those gradings to the number and let them drive decisions while the decisions are still available: reprice or exit thin work at renewal rather than at closeout, de-risk concentration before the break clause opens rather than after, and make capacity decisions ahead of a burn cliff instead of inside one. A break clause a few months out is a decision window with a closing date, and it belongs in front of the people who can act on it. A backlog treated this way stops being the number a firm hides behind and becomes the earliest place its next year can be read, and steered.