Pipeline, Backlog and Revenue Answer Three Different Questions

Most management packs contain revenue. Many contain a pipeline report. Very few contain backlog — and backlog is usually the number that would have answered the question being asked in the meeting.

Three numbers, three decisions

They're often used interchangeably in conversation, which is where the trouble starts.

Pipeline is work you are pursuing but have not won. It's a forward indicator with a probability attached. Useful for deciding whether to invest in business development, whether to keep an estimator, and whether the market is softening. Its horizon is roughly two to four quarters out, and it should always be weighted — an unweighted pipeline is a wish list.

Backlog is work you have won and contracted but have not yet delivered. It's not a probability; it's an obligation. This is the number that tells you how much capacity is already committed, what revenue is already secured, and how long the business can operate at current staffing before it needs new work. Its horizon is one to three quarters.

Revenue is work delivered and recognised. It's a historical record. It tells you what happened and, with margin attached, how well it happened. It tells you nothing about what's coming.

Each answers a different question. Substituting one for another produces confident decisions built on the wrong input.

The two failure patterns

The first is running the business on pipeline. This tends to happen in businesses with a strong sales culture or an optimistic founder. The pipeline looks healthy, so hiring proceeds, overheads increase, and capacity is built for work that hasn't converted. When conversion runs below assumption — which it does, routinely — the business is carrying cost against revenue that never arrived.

The second is running the business on revenue. This is more common in delivery-led operational businesses, and it's the more dangerous of the two, because revenue is a lagging indicator. By the time monthly revenue drops, the cause occurred three to six months earlier at the bidding stage. The business sees a resourcing problem four weeks before it lands, which is not enough time to either win work or reduce cost.

Backlog is the number that prevents both. It's known, it's contracted, and it's forward-looking.

Backlog needs assumptions too

Backlog isn't simply the sum of unbilled contract values. Two adjustments make it useful.

The first is a burn rate — how quickly the backlog converts to delivered revenue. A $9m backlog burning at $700,000 a month is fourteen months of work. The same backlog burning at $1.4m a month is six months. The dollar figure alone doesn't tell you which situation you're in.

The second is a margin overlay. Backlog at 22% gross margin and backlog at 11% are not equivalent, and a business can grow its backlog while reducing its future profitability — which is exactly what happens when tendering gets aggressive in a soft market.

Backlog reported without burn rate and margin is a vanity number.

One page, weekly

The practical version of this is a single page reviewed weekly:

• Weighted pipeline, with conversion assumption stated

• Backlog in dollars, in months of capacity, and at average margin

• Revenue month to date and rolling twelve months

• Delivery capacity available over the next two quarters

Four lines. Once those sit next to each other, the difficult conversations become straightforward.

Should we hire? Look at backlog months against current capacity, not pipeline.

Should we bid aggressively on this one? Look at backlog margin — if it's already compressed, another low-margin win makes the problem worse, not better.

Can we fund growth? Look at backlog and the working capital it will consume.

Should we decline this job? Much easier when you can see there are seven months of contracted work behind it.

The point

Most operators have a good instinct for where the business stands. Instinct is faster than reporting, and it's usually right, until it isn't.

What these three numbers provide isn't better instinct — it's the ability to test it, in about ninety seconds, before committing to a decision that takes two years to reverse.