How to Build a Lead Generation Strategy

Work backwards, always

Most lead generation plans start from last year's budget plus a percentage, then justify the number afterwards. A strategy starts from the revenue target and derives everything from it, which makes it falsifiable before you spend anything.

1. 01
   New revenue target ÷ average deal size = deals needed

   Use the median deal size in the segment you actually sell to, not the blended average across all segments, which is usually distorted by a handful of large deals.

2. 02
   Deals × coverage ratio = pipeline value required

   Coverage of 3× to 4× is typical. Use your own historical ratio if you have one — it is more reliable than a benchmark.

3. 03
   Pipeline value ÷ deal size = opportunities needed

   This is the number that matters most, because it is what channel performance is ultimately measured against.

4. 04
   Opportunities ÷ lead-to-opportunity rate = leads needed

   Use last year's actual rate, not an aspirational one. Planning on a conversion rate you have never achieved guarantees a miss you will attribute to volume.

5. 05
   Apply the timing lag

   Leads generated in Q1 close in Q2 or Q3. Budget against the quarter pipeline is needed, not the quarter revenue lands, or you will be permanently one cycle behind.

Derive your allowable cost

Before choosing channels, establish what a qualified opportunity is allowed to cost. Lifetime value, times gross margin, divided by a target LTV-to-CAC ratio of around 3:1, gives allowable acquisition cost per customer. Divide by win rate for allowable cost per opportunity.

That single number turns channel selection from a debate into a filter. A company with a $12,000 ACV discovers field events cannot work at any efficiency; a company at $200,000 discovers a $900 outbound lead is inexpensive. The full calculation is in what B2B lead generation costs.

Choose the channel mix

Two dimensions decide it: what your contract value supports, and how fast you need to learn.

Then split the portfolio by feedback speed. You need at least one fast-feedback channel that tells you within weeks whether the targeting is right, and at least one compounding channel that will still be producing in two years. Funding only fast channels means never building an asset; funding only slow ones means learning nothing for three quarters.

Two channels, run properly

The most common execution failure is running five channels at 20% effort each. Every channel has a minimum viable investment below which it produces noise rather than a signal, and five channels below that threshold produce five ambiguous results.

• Paid search — needs enough spend and enough weeks to exit the learning phase and reach statistical significance on conversion.
• Content and organic — needs two to four quarters before the read is reliable at all.
• Outbound — needs enough contacts per segment to distinguish a message problem from a list problem.
• Events — needs a follow-up motion designed before the event, or the spend is entirely wasted.

Pick two, fund them properly, and add a third only when one of the first two is either working or definitively dead.

Write down what you will not do

The section that makes a strategy usable and the one almost nobody includes. A plan listing only what you will do cannot be used to decline anything, which means it will not survive the first mid-quarter request from a senior stakeholder.

• Channels you are explicitly not testing this year, and why the economics do not work.
• Segments you are not pursuing, even though they occasionally buy.
• Asset types you will not produce — the ebook nobody reads, the report nobody cites.
• The request category you will decline — for instance, no one-off campaigns for a single sales rep's territory.

Instrument before you spend

Three things must exist before the first dollar goes out, because none can be retrofitted: source, medium, and campaign written to the record at creation; the first high-intent page viewed captured on the record; and a rejection reason required when sales declines a lead.

Without the first, you cannot attribute anything. Without the second, you cannot distinguish a researcher from a buyer. Without the third, the quality conversation stays an argument for as long as the programme runs. The full sequence is in the lead generation process.

Review on a fixed cadence

The quarterly row is the one that matters. A portfolio that never loses a channel accumulates underperformers indefinitely, because killing a channel requires someone to admit a decision was wrong — which is exactly why it should be scheduled rather than left to judgement.