Lead Generation Strategy: The Order You Decide Things In

Lead Generation Strategy: The Order You Decide Things In

Ask ten B2B teams what their lead generation strategy is and eight will answer with a channel. We are doing LinkedIn. We are investing in content. We are standing up an outbound team. Each of those is a real decision, and each one belongs fourth or fifth in the order, arriving first because it is the only one that feels like action.

The cost of taking the decisions out of order is not that you pick the wrong channel. It is that when the results come in, you cannot tell which decision was wrong. A programme that produces four meetings a quarter might have a targeting problem, an offer problem, a message problem, a channel problem or a volume problem, and if all five were decided at once, the only honest reading of four meetings is that something is wrong somewhere.

A strategy that can be debugged decides things in an order where each layer constrains the next, and where a bad result points at a specific layer.

The five layers, in the order they have to be decided

1. Step 11. Audience

   Which companies, and which person inside them. Narrow enough that you can name twenty examples.

2. Step 22. Offer

   The specific thing you are asking them to do, and why it is worth their time this month.

3. Step 33. Motion

   Buying meetings now, or lowering the cost of meetings later. These are different jobs.

4. Step 44. Channel

   Which route reaches that person, given the motion you chose.

5. Step 55. Volume

   How many attempts per period, set by what it takes to read a result.

The order is not arbitrary. Your audience decides which offers are credible. Your offer decides whether a fast motion is even available, because a low-commitment ask can be made cold and a high-commitment ask usually cannot. Your motion narrows the channel set to a handful. Only then does volume become a question with an answer, because volume is a function of the channel's cost per attempt and the conversion rate you are trying to measure.

Reverse it, and the constraints run backwards. A team that picks LinkedIn first ends up choosing an audience that happens to be reachable on LinkedIn and an offer that fits a LinkedIn message, which is a strategy assembled by a tool's affordances rather than by anything about the market.

Layer one, and the test that makes it real

Almost every weak strategy is weak here, and the tell is uniform: the audience is described in a way that cannot be falsified.

"Mid-market SaaS companies" cannot be wrong. "Series B to C horizontal SaaS companies between 80 and 400 employees who have hired a first RevOps person in the last year" can be wrong, which is precisely what makes it useful. If that definition produces nothing after a real attempt, you have learned something specific and you can change one clause of it.

The test to apply: can you name twenty real companies that fit, from memory or from ten minutes of looking, and can you say why the twenty-first was excluded? If the second half of that is hard, the definition is a description rather than a filter, and every number downstream of it will be uninterpretable.

There is a stronger version of the same move, which is defining the audience by something that happened rather than by something that is true. A company that has been mid-market SaaS for six years is mid-market SaaS again next quarter. A company that posted a specific role last month is a different list next month, and it contains a reason to write this week. The difference between a filter and a signal is the single highest-leverage thing at this layer, and it is worth resolving before anything below it is decided.

Layer two: the offer is the thing you are actually testing

The offer is what you ask for, not what you sell. A demo, a fifteen-minute call, an audit, a benchmark, an introduction. It is the decision most strategies skip entirely, defaulting to whatever the calendar link says.

Two questions decide it. What does this person get from saying yes, described in their terms rather than yours. And what does saying yes cost them, in time and in internal exposure. A senior person agreeing to a meeting with a vendor they have never heard of is spending something real, and an offer that ignores that cost is why polite non-answers dominate most reply sets.

The reason this layer sits above channel: an offer that only works after trust has been built rules out every cold channel, and discovering that after you have hired two SDRs is expensive.

Layer three: which job you are doing

Two jobs, genuinely different, and conflating them is the most common strategic error after the audience one.

Buying meetings now means paying per conversation, through outreach or paid or events, with results in days and a cost that does not fall much over time. Lowering the cost of meetings later means building something that makes your market arrive already knowing who you are, with results in quarters and a cost per meeting that falls.

Most companies need both eventually and can only afford to start one well. The question is which problem you have this quarter, and it is answered by the calendar rather than by preference. Sorting channels by how fast they answer is the practical version of this decision.

Layer four and five: channel, then volume, and why volume is the one that gets fudged

By the time you reach channel, most of the choice has been made for you, which is the point of the ordering.

Volume is where strategies quietly become unreadable. The number of attempts you need is not set by ambition, it is set by the conversion rate you are trying to detect. A channel converting somewhere near two percent produces one meeting per fifty attempts. Fifty attempts a month is one meeting a month, and one meeting a month is a number from which nothing can be concluded, in either direction, for a very long time.

Is this strategy debuggable?

• ✓Yes: The audience definition can be wrong, and you can say what would prove it wrong
• ✓Yes: The offer names what the prospect gets, not what you sell
• ✓Yes: You can say which of the two jobs this programme is doing
• ✓Yes: Volume is high enough that a monthly number means something
• ✓Yes: Only one layer is being changed at a time once it is running
• ✕No: The channel was decided before the audience
• ✕No: Three channels started in the same month

So volume is a gate on whether the strategy can be evaluated at all, not a dial to turn later. If your capacity cannot support enough attempts in your chosen channel for a monthly number to be stable, you have three options: narrow the audience so the same effort concentrates, change to a cheaper-per-attempt channel, or accept that you will be reading quarterly rather than monthly and stop having the conversation in between.

Changing one thing at a time, and how long to wait

Once it runs, the discipline that matters is boring and almost nobody keeps it: change one layer at a time, and wait until the current version has produced enough volume to be read.

A team that changes the list and the message in the same week has learned nothing by the end of it. This sounds obvious and it survives contact with reality for about three weeks, because a slow month creates pressure to change everything at once.

What helps is deciding the read date in advance, at the same time as the change. Not a number to hit, a date at which you will look. That converts a stressful ongoing judgment into a scheduled one, and it is the mechanism that keeps the one-thing-at-a-time rule alive.

Our own contribution to that discipline sits at the message layer: a campaign carries one message, built on one premise, and it is sent once. A later approach to the same person is a separate campaign with a separate premise and a separate read. That constraint makes the premise itself the variable under test, rather than letting repeated contact blur which premise earned the reply, and it means every result is attributable to a specific idea about why this person should care.

What the layers look like when they are wrong

A quick diagnostic, since the value of the ordering is that a bad result points somewhere specific.

Very low reply rates against an accurate list usually means the offer or the premise, not the audience. The list found real people at real companies and none of them thought the thing being asked was worth answering, which is information about the ask.

Good reply rates and no meetings usually means the offer is misaligned with what the replier actually wanted. People are engaging and then declining the specific next step, so the next step is the variable.

Meetings that go nowhere usually means the audience is one layer off, and most often it is the right company and the wrong person: someone senior enough to take the call and not the one who owns the problem. This is the failure that flatters itself, because the meeting count looks fine for a quarter before the pipeline does not.

A result that swings wildly month to month usually means volume rather than any of the above, which is why volume is a gate rather than a dial. Before diagnosing anything else, check that the monthly number is built on enough attempts to be a number at all.

And a programme that was healthy and decayed usually means the audience got used up rather than that anything got worse. A finite list contacted at a steady rate runs out, and the symptom is a slow decline that reads exactly like message fatigue. The distinction is checkable: count how much of the defined audience has already been contacted before rewriting anything. If the answer is most of it, the fix is at layer one, and no amount of message work will do it. Where the whole programme has decayed at once rather than in one layer, four frameworks for a broken lead generation system is the wider diagnostic.

The short version

Decide the audience, then the offer, then which job the programme is doing, then the channel, then the volume. Each layer constrains the next, so a result you do not like points at a layer rather than at everything. Define the audience so it can be falsified, set volume high enough that a monthly number is stable, change one layer at a time, and fix the read date before you make the change.

If the fast half of the strategy is the part you want built and running against a named list, we can put a campaign in front of your market and you can read it on the date you set.