Ask anyone what they're closing in six months

Ask an agency owner what they'll close over the next six months and you'll get a number. Ask how they got to it and you'll get a pause.

The number is usually last year's good quarter with a bit of optimism on top.

My car tells me I've got 210 miles of range (I’ve got EV, because I like the environment, and I like tax breaks). It does not have 210 miles of range. It has 210 miles if the road is flat, the heating is off and nobody else gets in. The number isn't a lie. It just assumes a day I'm not having.

Most pipelines are the same.

You open your pipeline. You can see what percentage of it will close and roughly when. You know the average contract value, because you've got enough history to know. So you can say, without crossing your fingers, that in six months you'll be doing this much work for this many clients.

Then you can decide the things that matter. Who to hire. When. Whether to take the bigger office. Whether to turn down the client you already know is going to ruin your Fridays.

That's predictable pipeline. Not more leads. The ability to make a promise about next March and keep it.

Very few businesses can do it. Before you read the next bit, be honest with yourself:

Gartner says 7% of sales organisations hit 90% forecast accuracy or better.¹ The B2B average is around 63%.²

XANT Labs went through 270,912 closed-won deals worth $18.1bn. Nearly half of them missed their forecast by more than half. The average ninety day prediction was out by over 31%.³ CSO Insights found nearly 60% of forecast deals slip into the next quarter.⁴

The maths underneath has got harder too. Median B2B win rates are down to about 19%, from 23% two years earlier.¹ B2B SaaS runs 18% to 25% from discovery to closed won.⁵ Which makes the 3x pipeline coverage rule wrong for most people, because 3x assumes you win one in three.⁶ At 20% you need 5x. Enterprise needs nearer 6x.

So for the average SaaS business: you can see about a quarter ahead, you'll be out by a third, and six in ten of the deals you're counting will arrive later than you said. If they arrive.

Agencies are no better. 93% of marketing and professional services firms say their growth engine isn't strong enough.⁷ Retainer relationships churn at about 18% a year and last around 56 months. Project work churns at 42% and lasts 24.⁸ Retainer revenue is three to four times more predictable.⁹ Project agencies start from nothing every January.

Nobody is seeing six months out. Most can't see to the end of the month.

Because you're not forecasting your own activity. You're forecasting someone else's timing.

Ehrenberg-Bass have the number on this. About 95% of your potential buyers are not in the market right now. Businesses change their main agency or platform every few years, so 20% are in play across a year and roughly 5% in any given quarter.¹⁰

If your marketing only points at the people with their hand up, your pipeline is a bet on when a stranger's contract runs out. Every competitor is bidding for the same 5%. That's why it costs what it costs, and it's why you can't forecast it.

We run the one of our clients newsletters. Around 12,000 views a month.

The engine is simple. Every couple of weeks we sit down with someone senior for twenty minutes and talk about whatever's live in ecommerce. That conversation becomes two articles a week, Tuesday and Thursday. It's run for over a year without a gap.

The readers are tech, finance and brand leaders at ecommerce businesses. Hardly any of them are ready to buy on the day it lands. That isn't the job.

What matters is what it does to the pipeline.

The audience is engaged and it's been fed for a long time. So when they want revenue in a given month, they don't go looking for new people. They go to the people already reading. Some outreach, some nurture, a tactical piece with a hard sell in it, pointed at the engaged segment. It works, near enough every time.

Which changes the question. It's no longer how do we fill Q4. It's when can we onboard the next one. That's a capacity conversation, and it's a much better one to be having.

We run a lot of research. Surveys, reports, benchmarks. With Rotate it's the Tech Leaders Survey, asking ecommerce CTOs and technical directors what they're planning for and what's getting in the way. The first one brought in over a hundred mid-market leads, none of them on Shopify, which was the exact list Rotate wanted (and Shopify).

The launch is the bit everyone gets excited about. It's the least interesting part.

Those leads didn't go straight to sales. They went into a six to eight week nurture sequence first, because a hundred people downloading a report is not a hundred people wanting a call. By the time they reached the sales process they knew who Rotate were.

Then the useful bit. Nearly twelve months on, that report is still producing leads every month.

That's on purpose. We don't do big bang. We take one piece of research and taper it out over twelve to twenty four months. Articles, social, ads, talks, follow-up angles, a second wave when the data gets refreshed. The launch spike is nice and it's short. The tail is where the predictability lives.

Once the tail is steady you can do maths on it. Spend this on ads, get this many MQLs. This many MQLs turns into this many SQLs. That's a model rather than a guess, and it's the closest thing to seeing six months out that I've found.

One caveat, and it's the whole thing. It only works if the research is worth reading on its own. The moment it reads like a sales process with a cover on it, the tail dies and you're back to fighting over the 5% with everyone else.

I play in a band. We rehearse whether or not there's a gig in the diary. Most weeks nothing comes of it.

But when someone asks if we can play in a fortnight, the answer is yes, because the set already exists.

A band that only rehearses once a gig is booked ends up turning gigs down. Not because they can't play. Because they can't play yet.

An audience works the same way. You feed it when nothing is happening, so something can happen at short notice.

Own the audience. Not followers on a platform that changes its rules every eighteen months. A list you can name, segment and email tomorrow morning.

Publish until it's boring. Two articles a week for a year beats a brilliant campaign in March. The rhythm is what turns reach into trust.

Taper, don't launch. One good piece of research should still be feeding you two years later. Cut it up, republish it, run it as ads, talk about it on stage. Launch week is the smallest part of it.

Know who's engaged. Opens, clicks, replies, who came to the event, who forwarded it internally. Most of your real pipeline is sitting in your ESP, not your CRM.

Keep a lever you rarely pull. The hard sell only converts because you've spent months not selling. Pull it every week and it stops working. Pull it quarterly and it's a switch.

Forget forecast accuracy. It's a lagging number and it mostly measures how honest people are being in the CRM.

The test is which question you're asking.

If it's still where the next client is coming from, you haven't got a pipeline.

If it's when you've got room to take them on, you have.