Stage 2: Outbound & qualification (with 2a PoC-led outreach)

Detail for stage 2 of e2e-stages. This is our constraint, so it has to run constantly. The biz dev group (Ola, Daniela, João, Monika, Luccas) works it every week on the Friday biz dev day, from 9 Oct 2026. Daniela is the proposed champion. Actions: beads epic da-3y4 (bd show da-3y4).

Output of the stage: qualified first conversations with high-value targets, handed into stage 3 (proposal, quote & RFP). The work has to be a steady weekly flow, not a burst.

The flow

  ┌──────────────────────┐
  │ 2.1 Pick targets     │  who is worth pursuing?
  └──────────┬───────────┘
             ▼
  ┌──────────────────────┐
  │ 2.2 Research account │  their portal, data, people, triggers
  └──────────┬───────────┘
             ▼
  ┌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┐
  ╎ 2a  Build a PoC      ╎  when worth it: a demo from their own public data
  └╌╌╌╌╌╌╌╌╌┬╌╌╌╌╌╌╌╌╌╌╌╌┘
             ▼
  ┌──────────────────────┐
  │ 2.3 Reach out        │  personal message (with PoC link), follow-ups
  └──────────┬───────────┘
             ▼
  ┌──────────────────────┐
  │ 2.4 First            │
  │     conversation &   │──── not a fit ───▶ drop, or park with a date
  │     qualify          │
  └──────────┬───────────┘
             ▼ qualified
     stage 3: proposal, quote & RFP

  running under all of it:
  2.5 Pipeline hygiene: every target in the CRM with a next step and date,
      reviewed at the Friday sync

Sub-steps: what "owns it" looks like

"Owns it" is level 2 in e2e-stages: done alone, on real targets, to our standard.

sub-step	owns it (level 2) when they…	evidence	judgement that stays human
2.1 Pick targets	choose targets against the ICP, portal first (see below), and can say in one line why each is worth it now	target list entries with a "why now" line	who is worth pursuing; positioning against incumbents
2.2 Research account	produce an account brief: current portal and stack, what data they publish, key people, procurement route, a trigger (tender, strategy, complaint, deadline)	brief in the CRM or account folder	what the real problem is
2a Build a PoC	build, alone and within the time box, a working demo from the target's own public data that shows something specific to them	live demo URL, time spent	whether a PoC is worth it for this target; avoiding competent-but-generic output
2.3 Reach out	send a message tailored to the target (not cosmetic personalisation), to a named person, and follow up to a set sequence	sent messages and follow-ups, reply or no reply	tone; what to lead with
2.4 First conversation & qualify	run the first call alone, then qualify or drop: need, budget or buying route, decision-maker, timing, fit with the catalogue	call notes plus a qualify/drop decision with its reason	whether to pursue; when to walk away
2.5 Pipeline hygiene	keep every target they own in the CRM with a current stage, a next step and a date	CRM state at the Friday sync	when to stop chasing

2.1 in more detail: portal-first targeting

We have a CRM (Twenty) and a shortlist, but no shared target list yet.

Work back from portals, not from company lists.

Find data portals worth improving: existing CKAN and other portals, e.g. registries such as dataportals.org, CKAN instance lists, and national open data catalogues.
Score them: size, how dated they are, the likely owner's budget, and any trigger.
Work back to a named contact at the owning organisation. This can be real work. AI prospecting tools (e.g. Apollo, Explee; not yet chosen) help find people and emails.
Load them into Twenty with an owner, a stage and a next step.

The result becomes the shared target list: one list, with each target having one owner.

2a in more detail: the PoC

A PoC used to be a sales deliverable. It is now cheap enough to build before we have spoken to anyone, and it signals interest and commitment far better than a briefing document.

Is it worth it? Only for targets past 2.1 and 2.2 with public data we can use. Otherwise it is free work nobody asked for.
Get the data from their existing portal, CKAN API or open data files. Check the licence allows reuse.
Build with the PortalJS skills inside the time box: half a day (agreed 2026-09-28).
Make it specific. It should show an improvement on their current portal: better search, visualisation, DCAT, speed or design. A generic demo sells nothing.
Host it on a demo URL, labelled as an unofficial demo by Datopian, with no implied endorsement.
Hand it to 2.3: "We looked at your portal, and here's what we've already built for you."

Evidence: demo URL, time spent, the outreach message that used it, and the response.

Levels for the stage as a whole

The numbers are suggestions, not targets yet; calibrate them after the first month of Fridays. Counted over a rolling eight weeks.

level	means	suggested markers
0 Not yet	Has not done outbound on a real target	—
1 With help	Does parts of the flow with someone else choosing targets, reviewing messages or running the call	contributes to ≥3 targets; e.g. builds a PoC someone else sends
2 Owns it	Runs the whole flow alone for their own targets	≥8 targets researched, ≥2 PoCs sent, ≥1 first conversation run alone, ≥1 qualify/drop decision; CRM current at every sync
3 Improves it	Owns it, and makes it easier for others	ships a reusable piece: a research skill, PoC template, sequence, ICP update or qualification checklist, used by someone else

Group flow: what we watch every Friday

Stage 2 is the constraint, so the group's flow matters more than any one person's level. At the Friday sync, the champion reviews:

this week	per person	group
targets researched
PoCs built
messages sent (new and follow-up)
replies
first conversations
qualified into stage 3

The question each week: is anything getting stuck, and where? For example, many PoCs but few replies points to the message or the targeting, not the PoC.

Starting points (first read, to confirm with each person)

person	likely start	first aim
Daniela	2 on 2a and 2.3 (Swiss partners)	level 2 across 2.1–2.5; as champion, move towards level 3 by building the kit
Monika	? (strong on client meetings)	2.3 and 2.4: first conversations and qualification
João	0–1 (builder)	2a: PoCs, then send them himself (2.3)
Luccas	0	2a: PoCs, then 2.3; his main development area (decided 2026-09-28)
Ola	0	2a with help, then alone; then 2.3

Kit: what exists, what to build

From training module 0. Status unknown; check at the first Friday sync.

kit	supports	status
ICP definition and target-list generation	2.1	shortlist exists; no shared list yet
AI prospecting tool (Apollo, Explee?) for contacts	2.1	to choose (action 3)
account research skill	2.2	?
PoC-led outreach skill (brief to demo, fast)	2a	?
outreach sequences with real personalisation	2.3	?
qualification checklist	2.4	not in training yet
CRM (Twenty) with the stages above	2.5	CRM exists; stages to set up

Guardrails

No prices and no promises in outreach. Pricing belongs to stage 3, against the catalogue.
PoCs are clearly labelled unofficial demos, and use data we are allowed to reuse.
Know each target's owner: one owner per target, so two people don't approach the same organisation.

Open questions

Which prospecting tool: Apollo, Explee, or something else? (action 3)
Who builds the first shared list from the shortlist plus portal-first search? A good first job for the champion.
Qualification: does inbound get the same treatment as outbound?