---
name: planning-demand-generation
description: >-
  Builds a quarterly B2B demand generation plan by deriving the real in-market
  percentage from interpurchase time, splitting budget between demand creation
  (brand) and demand capture (activation) against the 95:5 rule and the Binet
  and Field ratios, mapping programs to each bucket, and designing separate
  measurement for each. Use when the user says "demand gen plan", "marketing
  budget allocation", "brand vs performance", "pipeline plan for next quarter",
  "our MQLs don't close", "how much should we spend on brand", "always-on vs
  campaign", "channel mix", "we only do paid search", or "self-reported
  attribution". Do NOT use for defining the product's positioning canvas (see
  positioning-a-b2b-product), for category creation narrative work (see
  designing-category-narrative), or for testing brand asset recognition (see
  auditing-brand-distinctiveness).
metadata:
  version: "1.0"
---

# Planning demand generation

Produce one artifact: a quarterly demand generation plan with a budget split across demand creation and demand capture, a program list mapped to each bucket, and a measurement design that scores each bucket on different metrics. Out of scope: creative production, ad copy, sales territory or quota design, and the positioning inputs the plan depends on.

## Workflow

Copy this checklist into your reply and tick items as you complete them:

```
- [ ] 1. Collect inputs and the current effective split
- [ ] 2. Derive the in-market percentage from interpurchase time
- [ ] 3. Set the target brand/activation split and justify the deviation
- [ ] 4. Map every program to demand creation, capture, or conversion
- [ ] 5. Apply 70-20-10 within each bucket
- [ ] 6. Design measurement per bucket, including self-reported attribution
- [ ] 7. Run the validation loop until it passes
- [ ] 8. Emit the quarterly plan
```

**1. Collect inputs.** Ask for, or read from connected data: total quarterly marketing budget, the pipeline or revenue target, the current channel-by-channel spend, average sales cycle length, and average contract or renewal term. Then run a budget audit that maps every activity to brand or activation, including staff time and agency fees, not just media ([Growth Method summary of Binet & Field](https://growthmethod.com/long-and-short/)).

Compute and state the **current effective split** before proposing anything. Many teams discover their effective ratio is closer to 10:90 in favour of activation, and a 20:80 activation skew is common among performance-led teams and signals underinvestment in long-term growth ([Growth Method](https://growthmethod.com/long-and-short/)). Naming the current number is what makes the recommended shift arguable.

**2. Derive the in-market percentage.** Do not adopt a constant. Use the category's average interpurchase time; if it is unknown, survey buyers with "How frequently do you purchase X" using calibrated response categories such as "once in five years or less often / annually / each quarter" ([John Dawes, "The 95:5 Rule"](https://johndawes.info/the-955-rule/)).

Worked reference points from Dawes: a 5-year replacement cycle (a principal bank or law firm) implies about 20% in-market over a year and roughly 5% in a quarter; a 2-year cycle implies about 50% over a year and roughly 13% in a quarter ([Dawes](https://johndawes.info/the-955-rule/)).

State the derived number as an estimate with its assumption. The 95% figure is explicitly a heuristic, not a precise rule, and the true proportion varies by category, product, time period and interpurchase interval ([Dawes](https://johndawes.info/the-955-rule/)). Chris Walker's operating heuristic of 1-2% in-market at any time ([Refine Labs concepts write-up](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs)) is more aggressive than Dawes's calculation — cite it as a floor, and use the interpurchase-time derivation as the planning number.

**3. Set the target split.** Anchor on the evidence base: brands allocating roughly 60% to long-term brand building and 40% to short-term activation achieved the best long-run results across approximately 1,000 IPA Databank case studies — higher market share, stronger pricing power, lower cost of acquisition ([Growth Method](https://growthmethod.com/long-and-short/)). For B2B, the LinkedIn B2B Institute worked with Binet and Field in 2019 to recalibrate to approximately **46% brand / 54% activation** ([Growth Method](https://growthmethod.com/long-and-short/)). Use 46:54 as the B2B starting point.

Then adjust with stated reasoning, because 60:40 is a guideline and the average of a distribution, not a law ([Growth Method](https://growthmethod.com/long-and-short/)):

| Condition | Direction | Reason |
|---|---|---|
| Established, high awareness in category | More activation | Existing memory structures already do the priming work |
| New entrant or unknown brand | More brand | Activation converts poorly with no mental availability to draw on |
| High-frequency / short-trial purchase | More activation | Buying windows open often, so capture has more to catch |
| Low-frequency, high-consideration (enterprise software, professional services) | More brand | The buying window is narrow and infrequent, so memory must be built in advance |

Write the recommended split as a single number pair with a one-sentence justification naming which row applied. Never present the split as a rule handed down; present it as a diagnostic against the current effective split from step 1.

**4. Map every program to a phase.** Use the three demand phases and assign each planned program to exactly one ([Refine Labs concepts](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs)):

| Phase | Job | Typical programs |
|---|---|---|
| Demand creation | Educate and build preference in the out-of-market majority | Podcast, paid social, organic social, community evangelism, third-party events |
| Demand capture | Convert existing in-market intent | SEO/SEM, review sites, affiliates, website CRO, retargeting, SDR outbound |
| Demand conversion | Close | AEs, solutions consultants, sales engineers |

Demand creation maps to the brand side of the split, demand capture to activation. Paid search and retargeting are inherently activation; video, PR, podcast and organic SEO build brand over longer timeframes ([Growth Method](https://growthmethod.com/long-and-short/)). If a program plausibly sits in two phases, force a single assignment and note the ambiguity — dual-counting is how activation-heavy plans disguise themselves as balanced.

Flag any plan where creation spend is zero. If a company only captures and converts, the addressable audience shrinks over time and the company battles over the small in-market slice indefinitely ([Refine Labs concepts](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs)).

**5. Apply 70-20-10 within each bucket.** Split each bucket's spend 70% proven channels, 20% emerging, 10% experimental, and keep the 70% core itself internally balanced brand versus activation ([Growth Method](https://growthmethod.com/long-and-short/)). This is a default, not a constraint — if the user has a hard reason to deviate (a single channel carrying the business, a mandated pilot), adapt and record the reason in the plan.

**6. Design measurement per bucket.** Measure each bucket differently: activation on short-window conversion metrics; brand on awareness, consideration and price premium, never on last-click ([Growth Method](https://growthmethod.com/long-and-short/)). Judging creation programs on last-touch conversion is the single most common way a good plan gets cancelled in month two.

For brand-side measurement, track mental availability through links to **category entry points** — the situations in which the category could be bought or used ([Dawes](https://johndawes.info/the-955-rule/)). Set expectations against the benchmark: building double-digit mental availability is a multi-year task, many well-established brands reach no higher than 20-30% of respondents linking them to a category entry point, and even market leaders often reach only 50% ([Dawes](https://johndawes.info/the-955-rule/)).

Specify the self-reported attribution instrument exactly, because the implementation details determine whether the data is usable. Add a **mandatory free-text field** on the primary declared-intent conversion form asking **"How did you hear about us?"** with **no leading text and no suggested options**, and run it for **30 days** before judging it ([Refine Labs concepts](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs); [Megan Bowen, Refine Labs](https://www.refinelabs.com/blog/attribution-mirage)). Do not replace it with a dropdown; suggested options bias the answer set toward the channels already believed to work.

Read the two systems as answering different questions: self-reported tells you where people first hear about a vendor, software attribution tells you where people first interact digitally with a vendor ([Refine Labs](https://www.refinelabs.com/blog/attribution-mirage)). Expect divergence — in Refine Labs' 12-month test across 620 conversions, software attribution credited 78% of conversions to direct plus organic search while customers self-reported 12%, and 85% of self-reported conversions and 98% of closed-won revenue traced to dark social, with podcast and community not measured at all by the software ([Refine Labs](https://www.refinelabs.com/blog/attribution-mirage)). State plainly that this sample is Refine Labs' own funnel — an agency whose audience skews podcast and LinkedIn-native — so the direction is replicated but the specific 85% and 98% figures are not universal benchmarks.

**7. Validate, fix, re-validate.** Run every check. If any fails, fix the plan and re-run the full list. Only emit the plan when all checks pass.

```
- [ ] Current effective split is computed and stated, not assumed
- [ ] In-market % is derived from interpurchase time, not copied from 5% or 1-2%
- [ ] Target split names its starting anchor (B2B 46:54) and its adjustment reason
- [ ] Every program is assigned to exactly one phase, no dual counting
- [ ] Demand creation spend > 0
- [ ] Bucket spend sums to the stated total budget
- [ ] No creation program is scored on a last-click or short-window metric
- [ ] Self-reported attribution field is free text, mandatory, no options, 30-day run
- [ ] Every cited benchmark carries its source link
- [ ] Spend is spread across the quarter, not concentrated in one burst
```

The last check matters because there is a hard practical ceiling on how many buyers any single campaign can acquire in a period, and concentrating budget in one window leaves the brand off-air while later buyers enter the market ([Dawes](https://johndawes.info/the-955-rule/)).

## Output format

Use this exact section order and headings; the split table and program table are read positionally by downstream reporting. Numbers and rationale inside are yours to adapt.

```markdown
# Demand generation plan — <company>, <quarter>

## Market timing
- Interpurchase cycle: <n> years (source: <how derived>)
- Estimated in-market this quarter: <n>% — the remaining <n>% must be educated, not converted
- Implication: <one sentence>

## Budget split
| Bucket | Current effective | Recommended | Delta | Reason |
|---|---|---|---|---|
| Demand creation (brand) | <n>% | <n>% | <+/-n> | <row applied from step 3> |
| Demand capture (activation) | <n>% | <n>% | <+/-n> | |

Anchor: B2B 46:54 brand/activation. Deviation justified by: <condition>.

## Programs
| Program | Phase | Spend | 70-20-10 tier | Owner | Primary metric |
|---|---|---|---|---|---|

## Measurement design
| Bucket | Metrics | Cadence | Explicitly NOT judged on |
|---|---|---|---|
| Creation | | | last-click, MQL volume |
| Capture | | | |

**Self-reported attribution:** mandatory free-text "How did you hear about us?" on <form name>, no options, review after 30 days.

## Risks and assumptions
- <assumption> — invalidated if <observable signal>
```

## Freedom calibration

- **Fixed:** deriving the in-market percentage rather than assuming it, one-phase-only program assignment, separate metrics per bucket, and the exact self-reported attribution field wording. These are the steps where a plausible-looking shortcut produces a plan that measures the wrong thing.
- **Your judgment:** the final split, which programs to fund, how to weight the 70-20-10 tiers, and how to sequence spend across weeks. State reasoning; do not seek a formula.
- **Ask the user:** budget totals, current spend by channel, sales cycle length, and contract term. Do not estimate these from category averages, because the entire in-market derivation hangs on the cycle length.

## Gotchas

- Advertising cannot principally work by stimulating immediate purchase, because most of it reaches people who will not buy soon; it works by building and refreshing memory links activated later when the buyer enters the market ([Dawes](https://johndawes.info/the-955-rule/)). Any plan whose creation programs are held to in-quarter conversion targets is internally inconsistent, and the inconsistency should be named in the risks section.
- Targeting only in-market searchers underperforms even for capture, because people largely use memory when buying rather than searching, and when they do search they strongly prefer familiar brands — lesser-known brands show lower consideration rates and materially lower click-through rates ([Dawes](https://johndawes.info/the-955-rule/)). Brand spend therefore lowers the cost of the capture programs it sits alongside.
- Most companies fund and measure only capture and conversion, which creates a divide within the revenue team and means the metrics incentivize marketers to capture demand even when leaders ask for creation ([Refine Labs concepts](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs)). Fix the metric before arguing about the budget; a creation budget measured on MQLs reverts within a quarter.
- Software attribution dramatically over-reports organic search and direct traffic and significantly under-reports social, and it cannot see LinkedIn and Slack conversations, podcasts, communities, DMs, or meet-ups at all ([Refine Labs](https://www.refinelabs.com/blog/attribution-mirage)). A dashboard showing SEO as the top source is usually showing where dark-social-influenced buyers typed the brand name they already knew.
- Separate earned demand (demand the company creates) from organic demand (word of mouth and other things outside the company's control) in reporting; most traffic in attribution reports should come from earned ([Refine Labs concepts](https://www.anoopaulakh.com/blog/full-funnel-attribution-and-demand-generation-core-concepts-from-refine-labs)). A plan that quietly banks on organic demand is not a plan.
- The Binet and Field dataset is predominantly B2C consumer goods, which is precisely why the 46:54 B2B recalibration exists ([Growth Method](https://growthmethod.com/long-and-short/)). If a stakeholder quotes 60:40 at a B2B company, correct to 46:54 rather than arguing the underlying evidence.