The demand generation system, in twelve stages.

Stage 1 of 12 · Know who to sell to

1. Know who to sell to

   Ideal customer profile (ICP)

   A written description of the companies most likely to buy: what they do, how big they are, what they run. Sales and marketing work from the same definition.

   Firmographics, technographics and the operational signals that predict need, written as a definition with priority tiers and mapped to industry codes. Sales signs off before a campaign runs.

   Where it breaks. When the ICP lives in people's heads, marketing and sales build two different lists.

   How I built one from federal data

2. Know who to sell to

   Multichannel top of funnel

   Fill the top of the funnel from more than one place: the website, search, email, LinkedIn, trade shows, outbound and purchased lists. No single channel carries the number.

   Inbound from the site, search and content; paid and organic social; events with lead capture; outbound sequences; list sources. Each is tagged at entry, so every contact arrives with a channel, a campaign and a date.

   Where it breaks. When one channel carries most of the number, a slowdown in that channel becomes a slowdown in pipeline, and there is no second source to make up the difference.

3. Get the data right

   Dedupe and clean

   One record per company and one per person, with the same words meaning the same thing everywhere.

   Duplicate rules on domain and email, standardized picklists, required fields at each stage and a recurring exception report. Bad data is stopped at entry, not cleaned up later.

   Where it breaks. Every score, routing rule and dashboard downstream inherits the duplicates and inconsistencies in these records.

4. Get the data right

   Enrich

   Fill in what you don't know about each company and contact, so a rep never opens an empty record.

   Firmographic and contact enrichment from a data vendor on create and on a schedule, industry codes on the company record, and company data cascaded down to its contacts.

   Where it breaks. Enrichment that runs once is stale in a quarter. It needs a schedule and an owner.

5. Get the data right

   Grade the fit

   Grade every company on how well it fits the ideal customer profile: A, B, C or D. This is the firmographic half of the lead score.

   A firmographic score built from industry, size, geography, tier and the operational signals in the ICP, expressed as a grade the whole revenue team can read at a glance.

   Where it breaks. Without a grade, the most active lead gets worked ahead of the best-fitting account.

   How I built the scoring model

6. Get the data right

   Segment and prioritize

   Sort companies into groups by fit, industry and role, so the message matches the reader and the best accounts get worked first.

   Segments by grade, vertical and buying-committee role, stored as properties rather than spreadsheets, so campaigns, routing and reporting all use the same ones.

   Where it breaks. A segment that exists only in a planning deck cannot drive a workflow, a list or a report.

7. Create and work demand

   Outbound

   Reach out to the right companies on purpose, in a sequence, instead of waiting for them to find you.

   Sequenced calls, emails and social touches for the highest-grade accounts, personalized research for the top tier, and every touch logged as an activity the funnel can be measured on. Coverage is planned against seller capacity.

   Where it breaks. A team can properly work a fixed number of accounts. If the sequence is pointed at the wrong tier, the best accounts end up in automated nurture.

   How I matched coverage to capacity

8. Create and work demand

   Nurture

   Stay in touch with people who are interested but not ready, with content worth reading.

   Nurture tracks keyed to stage, segment and behavior; content by vertical; paid social to the ICP; re-engagement when activity returns.

   Where it breaks. Sending the same email to every contact is a newsletter. Nurture is timed and segmented by what the person has shown interest in.

9. Create and work demand

   Engagement score

   Score what each person does: visits, downloads, replies, conversations at events. Firmographic grade plus engagement score equals the lead score, and when it crosses the threshold for a marketing-qualified lead (MQL), the lead goes to sales.

   Behavioral points for web, email, content, event and intent activity, with decay on inactivity and a written MQL threshold. Lead score = firmographic score + engagement score, stored as a property so routing and reporting read the same number.

   Where it breaks. Without an MQL definition written down and agreed with sales, the score is a number nobody has to act on.

   How inbound got a 24-hour standard

10. Prove it

    Attribution

    Know which touches led to which deals, from the first visit to the closed opportunity.

    Every stage change timestamped, campaign touches tied to contacts and companies, marketing-sourced and marketing-influenced pipeline defined in writing, multi-touch where the data supports it.

    Where it breaks. Attribution stops at the form fill unless the CRM (the customer relationship management system) carries the campaign through to the opportunity.

11. Prove it

    Return on investment (ROI)

    Show what the marketing dollars returned in pipeline and revenue, by channel and by campaign.

    Spend by channel and campaign against sourced and influenced pipeline; cost per MQL, SQL (sales-qualified lead) and opportunity; payback by segment, reviewed on a cadence with sales leadership.

    Where it breaks. When marketing reports in leads, leadership asks for more leads. When it reports in pipeline and dollars, the conversation moves to what the team needs in order to produce more.

12. Prove it

    Reporting

    Dashboards for the executive, for each segment lead and for each rep, all reading the same numbers.

    An executive view of stage volume and conversion, segment views for the leaders who own a vertical, and rep views of assigned leads, activity and next actions, all built on the same properties.

    Where it breaks. When three dashboards use three definitions of a lead, every review starts with an argument about the numbers instead of a decision.

    How the whole system reported