Buying Group Map · October 2026

Wrenstall Automation Ltd

Designs, builds and installs robot cells, conveyors and line control systems for UK manufacturers, from a build shop in Redditch. 74 staff.

Worked example. Wrenstall Automation and its figures are fictional. All figures come from the worked-example spreadsheet: ALL CHECKS PASSED (with 6 notes).
The headline

60% of deals had one role met before the proposal, or none

Two or more roles met before the proposal
40%
of 45 closed deals; on average 3.3 roles involved, 1.6 met before the proposal
Win rate, decider met before the proposal
65% v 21%
17 deals against 24; on CRM and email records alone 85% v 36%. An association, not proof
Roles that need a card
10 of 10
each with the content to prepare, the meeting to ask for and the question to ask early, staged
Catalogue uses
9 fed
5 ready now, 4 partly ready, 0 waiting

Verdict: 45 closed deals across 2 segments won 36%. On average 3.3 roles were involved in a deal but we met 1.6 before the proposal; 60% of deals had one role met before it, or none. Where we met the decider before the proposal we won 65% (of 17), against 21% where we didn't (of 24): beyond chance; on CRM and email records alone it holds. An association, not proof: a meeting late in a deal can follow winning rather than cause it. Losses rest more on salespeople's recall than wins (52% against 0%), so read the comparisons on records only. The decider is most often Owner/MD (17); Finance (10) most often appears late. 11 role cards cover 10 of the 10 roles that need one, each with the content to prepare, the meeting to ask for and the question to ask early; 2 of 2 segments are ready to use. It feeds 8 catalogue products: of 9 uses, 5 are ready now and 4 partly ready (they need something else as well). Measure later: 20% of open deals have two or more roles met before the proposal now; by the re-measure it is borderline for a realistic change; the win rate can only show direction. Costs: £7,744 one-off and £11,138 a year of extra meetings. Measure again on 2 May 2027 (early read 31 Jan 2027).

Who is involved

Every role, across the 45 closed deals

Owner/MD18 deals · decided 17 · blocked 2 · late 3Finance28 deals · decided 11 · blocked 7 · late 10Operations45 deals · decided 11 · blocked 0 · late 0End user3 deals · decided 0 · blocked 0 · late 0Procurement20 deals · decided 2 · blocked 5 · late 7Technical/H&S32 deals · decided 0 · blocked 4 · late 5Other1 deal · decided 0 · blocked 0 · late 1

Met before the proposal Met later Involved, never met Each bar is the number of closed deals the role took part in, split by when someone from the business first met them.

Before this, salespeople worked through the operations or engineering manager who sent the enquiry. The deals show finance, technical and the owner in most of the larger ones, usually met late or not at all, and purchasing in most automotive deals, arriving at negotiation.

How the deals were recorded

Recall forgets the people we never met

33% of deals from recall: 0% of won deals, 52% of deals not won: a gap of 52 points, so read the comparisons on records only too.

Every won deal was rebuilt from the CRM and emails, while half the losses rest on salespeople’s memory, which tends to remember only the contact. So each comparison below is shown twice: on all deals, and on the 27 deals recorded from the CRM and emails alone.

What goes with winning

Meetings before the proposal: associations, not proof

ComparisonWithWithoutGap ± give or take (points)Result
Decider met before the proposal (decider known): all deals65% of 1721% of 2444 ± 28Beyond the give or take: an association, not proof
Decider met before the proposal (decider known): CRM and email records only85% of 1336% of 1449 ± 33Beyond the give or take: an association, not proof
Finance met before the proposal (finance involved): all deals100% of 328% of 25–Too few to compare (fewer than 5 deals in a group)
Finance met before the proposal (finance involved): CRM and email records only100% of 350% of 14–Too few to compare (fewer than 5 deals in a group)
Two or more roles met before the proposal, against one or none: all deals67% of 1815% of 2752 ± 26Beyond the give or take: an association, not proof
Two or more roles met before the proposal, against one or none: CRM and email records only71% of 1740% of 1031 ± 38Within the give or take: could be chance
New robot cell or line: decider met before the proposal (decider known)100% of 514% of 1486 ± 19Beyond the give or take: an association, not proof
New robot cell or line: two or more roles met before the proposal, against one or none55% of 1111% of 943 ± 37Beyond the give or take: an association, not proof
Upgrade or retrofit: decider met before the proposal (decider known)50% of 633% of 917 ± 52Within the give or take: could be chance
Upgrade or retrofit: two or more roles met before the proposal, against one or none86% of 70% of 1086 ± 26Beyond the give or take: an association, not proof
Controls and software: decider met before the proposal (decider known)50% of 60% of 1–Too few to compare (fewer than 5 deals in a group)
Controls and software: two or more roles met before the proposal, against one or none0% of 038% of 8–Too few to compare (fewer than 5 deals in a group)
Roles met before the proposalDealsWonWin rate
None00–
127415%
211545%
355100%
4 or more22100%

The comparisons count only meetings before the proposal: a meeting with the owner or finance late in a deal can follow winning rather than cause it, because customers who have decided introduce the people who sign. Even so, deals that were going well may be the ones where we got to meet more people early. The decider comparison holds on records alone; the two-or-more comparison doesn’t (it could be chance there), and finance met before the proposal is too rare to compare. Small groups, including most splits by deal type, are suppressed or should be read as hints.

When roles appear late

The stage a new role first appeared

StageDealsOf which not won
Enquiry00
Discovery or survey21
Proposal109
Negotiation119
Sign-off33

Decides: Owner/MD (17). Blocks: Finance (7). Appears late: Finance (10), most often at negotiation.

The map, by segment

Who is involved, who decides and blocks, and the deal checklist

S1 · Food and drink producers

24 closed deals, 9 won (38%). On average 3.1 roles involved and 1.6 met before the proposal; 33% of deals had two or more roles met before the proposal. Most often decides: Owner/MD (11). Most often blocks: Finance (4). Most often appears late: Finance (6).

RoleIn dealsMet when involvedDecidedBlockedLateUsually appears
Operations
Operations manager; production manager; site engineering manager
24 (100%)100%600Enquiry
Technical/H&S
Technical manager; food safety or hygiene manager; health and safety adviser
18 (75%)44%034Discovery or survey
Finance
Finance director; financial controller; group finance in larger groups
17 (71%)35%646Proposal
Owner/MD
Owner; managing director (family-owned producers)
12 (50%)50%1112Proposal
End user
Shift leaders; line operators; maintenance technicians
2 (8%)50%000Discovery or survey
Procurement
Purchasing manager in group-owned producers
2 (8%)50%011Sign-off

Deal checklist, S1

WhenDo thisAsk earlyBring
First meetingAsk who else will be involved in deciding and who signs off; record each role against the deal in the CRM, and each meeting with its dateWho else will be involved in deciding, and who signs off?–
By enquiryOperations (usually appears at enquiry): Trial or simulation review with the operations managerWho else will be involved in deciding, and who signs off spend of this size?Trial or simulation results on the customer's own product; the installation timetable
By discovery or surveyOwner/MD (usually appears at proposal): Reference-site visit, or a 30-minute call with our managing directorWill the owner want to meet us before deciding, and at what stage?Installation and fallback plan; two reference sites willing to host a visit
By discovery or surveyTechnical/H&S (usually appears at discovery or survey): Technical review at the site surveyWho signs off food safety and machinery safety, and can they join the survey?Hygiene and guarding pack: cleaning method, materials, risk assessment outline and the technical file contents list
By discovery or surveyEnd user (usually appears at discovery or survey): Ten minutes with a shift leader and a maintenance technician at the surveyWho will run and maintain it, and can we talk to them at the survey?Short video of changeover and jam clearing on a similar cell; training plan by shift
By discovery or surveyProcurement (usually appears at sign-off): Call with purchasing before the proposal to agree whose terms applyDoes the order go through a group purchasing team, and whose terms will it be on?Supplier approval pack: questionnaire answers, insurance certificates, accreditations and our usual positions on the terms
Gate: before the proposal goes outCheck that two or more roles have been met in a meeting, including the one that usually decides or blocks; if not, ask for that meeting before sendingWhat will the people who sign off need to see?–
By proposalFinance (usually appears at proposal): Payback review with finance before the proposal goes to the boardWho signs off spend of this size, and what does the board need to see?One-page payback sheet from the customer's own figures, with a half-as-well case; payment schedule

S2 · Automotive and engineered component makers

21 closed deals, 7 won (33%). On average 3.4 roles involved and 1.6 met before the proposal; 48% of deals had two or more roles met before the proposal. Most often decides: Owner/MD (6). Most often blocks: Procurement (4). Most often appears late: Procurement (6).

RoleIn dealsMet when involvedDecidedBlockedLateUsually appears
Operations
Operations manager; manufacturing engineering manager
21 (100%)100%500Enquiry
Procurement
Purchasing manager; group procurement; supplier quality
18 (86%)33%246Negotiation
Technical/H&S
Quality manager; manufacturing engineer; health and safety manager
14 (67%)64%011Discovery or survey
Finance
Group finance; finance director
11 (52%)27%534Proposal
Owner/MD
Managing director; site general manager
6 (29%)33%611Proposal

Deal checklist, S2

WhenDo thisAsk earlyBring
First meetingAsk who else will be involved in deciding and who signs off; record each role against the deal in the CRM, and each meeting with its dateWho else will be involved in deciding, and who signs off?–
By enquiryOperations (usually appears at enquiry): Trial review with the operations and quality managersWho else will be involved in deciding, and is there a capital approval form?Trial results on the customer's own parts, with cycle time and capability
By discovery or surveyProcurement (usually appears at negotiation): Call with purchasing before the proposal to agree whose terms applyAre we on your approved supplier list, and whose terms will the order be on?Supplier approval pack: questionnaire answers, insurance certificates, accreditations and our usual positions on the terms
By discovery or surveyTechnical/H&S (usually appears at discovery or survey): Quality and safety review at the trialWho approves quality and machinery safety on a new cell, and can they see the trial?Capability results and the documentation list; risk assessment outline and guarding approach
By discovery or surveyFinance (usually appears at proposal): Thirty minutes with finance to check the approval form before it goes inIs there a capital approval form and a payback limit, and who completes it?Capital approval form completed with the customer's own figures and a cautious case
By discovery or surveyOwner/MD (usually appears at proposal): Short meeting between the two managing directorsDoes the managing director sign this off, and when would they want to meet us?Start-of-production plan with milestones; two references from similar component makers
Gate: before the proposal goes outCheck that two or more roles have been met in a meeting, including the one that usually decides or blocks; if not, ask for that meeting before sendingWhat will the people who sign off need to see?–
Role cards

Roles and evidence only: what each fears, needs and asks

RoleWhat they fearEvidence they needWhat they ask
S1 · Operations
met in 24 deals; 4 interviews or notes
The installation running into production; the line not reaching rate; being blamed if it doesn't workA trial or simulation on their own product; the installation timetable; what rate we commit to and how it is testedHow long will the line be down? What rate will it really run at? Who supports it at three in the morning?
S1 · Technical/H&S
met in 8 deals; 3 interviews or notes
Equipment that can't be cleaned properly; a guarding or PUWER finding after installation; a non-conformance at the next auditHygienic design details and the cleaning method; the risk assessment and guarding approach; the declaration of conformity and technical file contents; references from other food sitesHow is it cleaned, and how long does it take? Who signs the risk assessment? What do we get at handover?
S1 · Finance
met in 6 deals; 3 interviews or notes
A project that overruns or pays back later than promised; committing cash before the order book supports itPayback built from the customer's own figures (labour hours, giveaway, downtime) with a cautious case; staged payments; what is and isn't includedWhat is the payback if output is only half what you say? What does it cost us if it's late? Can the payments be staged?
S1 · Owner/MD
met in 6 deals; 3 interviews or notes
Disrupting supply to a major retailer during installation; backing the wrong supplier; a project the staff won't acceptAn installation plan built around production with a fallback; references from similar owner-run producers; who supports it after handoverWhat happens if it doesn't work on day one? Who do I call? Have you done this for someone like us?
S1 · End user
met in 1 deal; 1 interview or note; thin evidence
Equipment that is hard to clear when it jams; more work at changeover; not being trainedChangeover and jam-clearing shown on a similar cell; a training plan by shift; the spares listHow do we clear a jam? How long is changeover? Who trains the night shift?
S1 · Procurement
met in 1 deal; 1 interview or note; thin evidence
A supplier that isn't approved; terms that leave the group exposed; quotations that can't be comparedThe supplier questionnaire completed; insurance certificates; our terms marked against theirs; a price breakdownAre you on our approved list? Will you accept our terms? Can you break the price down?
S2 · Operations
met in 21 deals; 3 interviews or notes
A cell that doesn't hold cycle time or quality; the changeover onto a new part taking weeksCycle-time and capability results from a trial on their own parts; the changeover plan for new partsWhat cycle time will it hold? How quickly can we change it over for a new part?
S2 · Procurement
met in 6 deals; 3 interviews or notes
A supplier that isn't approved; unlimited liability or warranty terms; prices that can't be compared like for likeThe supplier questionnaire completed; insurance certificates and accreditations; our terms marked against theirs; a price breakdownAre you on our approved list? Will you accept our terms? What are your payment terms?
S2 · Technical/H&S
met in 9 deals; 2 interviews or notes
A cell that passes the trial but drifts in production; missing quality documentation; a guarding finding after installationCapability results on their parts; the quality documentation the customer's own customers require; the risk assessment and guarding approachWhat capability did it show on our parts? What documentation do we get? Who signs off the guarding?
S2 · Finance
met in 3 deals; 2 interviews or notes
Payback beyond the group's limit; prices that change after approvalPayback in the group's own format from the customer's figures, with a cautious case; a fixed price and its validityWhat is the payback on our numbers? How long is the price fixed for?
S2 · Owner/MD
met in 2 deals; 2 interviews or notes
A cell that delays a new contract's start of production; a supplier the group hasn't heard ofThe start-of-production plan with its milestones; references from similar component makersWill it be ready for the start of production? Who else have you done this for?

No invented people: no names, ages or hobbies. Each card counts the deals in which the role was met plus the interviews and notes behind it, and lists them in the workbook. Two cards (food producers’ end users and purchasing) rest on thin evidence. Full cards, with how to reach each role and the actions, owners, stages and dates, are in the workbook and the pack.

How the value is shown

Unlocks: the catalogue products this feeds

ProductSegmentWhat it needsWhat its own kit asks forStatus
Proposal RewriteAllRole cards: evidence each role needsProof for the template: case studies with the client's permission, CVs, insurance certificates, standard terms (its SKILL.md)Partly ready: also needs the proof itself (case studies with permission and payback examples), made by the actions on the cards
Objection Handling LibraryAllRole cards: fears and questionsThe objections, with stage, deal reference and source; its defaults include "decision-maker not here" (its SKILL.md)Partly ready: also needs a two-week tally of the objections the team hears
Value Proposition and MessagingS1Role cards: who buys and what each cares aboutTarget segments (up to four) and who buys in each (its SKILL.md)Ready now
Value Proposition and MessagingS2Role cards: who buys and what each cares aboutTarget segments (up to four) and who buys in each (its SKILL.md)Ready now
Call and Appointment ScriptsAllChecklist: who to ask for and the early questionA conversation is a call that reached the decision-maker (its method)Ready now
Account Review PackAllRoles to keep in contact withWho you deal with at each account, by role, with the last date spoken to or met (its SKILL.md)Partly ready: also needs a contact log for each key account by role and date
CRM BlueprintAllBuying roles for the CRMA field dictionary with picklist values; its contact role picklist (its SKILL.md and worked example)Ready now
Stalled Deal RescueAllDecider and blocker patternsThe six answers for each open deal, including no decision-maker, no budget holder and no champion (its SKILL.md)Ready now
Win-Loss ReviewAllDecider and blocker patternsSix to ten customers (some who said yes, some who said no) willing to talk for about 15 minutes (its SKILL.md)Partly ready: also needs six to ten customers willing to be interviewed

Partly ready means this map supplies its part, but the product needs something else too: willing interviewees for the Win-Loss Review, a contact log for each key account for the Account Review Pack, the proof itself for the Proposal Rewrite, and a tally of objections for the Objection Handling Library.

Measure later

Open deals with two or more roles met before the proposal; the win rate as direction only

Main measure: 8 of 40 open deals (20%, give or take 13 points); spot-check of 10 deals: 16 of 17 logged roles confirmed, 1 met but not logged. With 40 open deals expected, the smallest change it can show (80% power, at 20% and 50%) is 28 points against a realistic 30 (assumed): borderline, so read the result with care.

Win rate: 36% now; with 22 deals closing by the re-measure, the smallest change it can show is 36 points: direction only.

Counted like for like: open deals created in the six months before each count and at least 30 days old, from the CRM; a role counts as met only in a meeting or site visit before the proposal, and at least one of the roles must be one that decides or blocks. A spot-check against diaries and email both confirms what the CRM says was met and finds what it missed, both times. At the re-measure only deals created after go-live count. Go-live, early read and re-measure: 2 Nov 2026; 31 Jan 2027; 2 May 2027.

Costs: £7,744 one-off and £11,138 a year of extra meetings. Illustration only: £95,693 a year in gross profit at 40%; £47,847 if it moves half as far (what it would be worth if the owner’s target win rate were reached is an illustration, not a forecast).

How confident are we?

What this rests on

The deals are 45 closed deals from the business's CRM, emails and quotation log, with the roles involved, the stage each was first met, who decided, who blocked and who appeared late as the business recorded or remembered them (33% from salespeople's recall, which forgets the people we never met: 0% of won deals and 52% of deals not won). The win rates side by side are associations, not proof: deals that were going well may be the ones where we got to meet more people, and a meeting late in a deal can follow winning rather than cause it, so the comparisons use meetings before the proposal and are shown on CRM and email records alone too; a comparison is shown only with at least 5 deals in each group. The role cards are the team's judgement from the deals where each role was met and 27 interview and note references, each card listing its evidence; which roles need a card (20% of a segment's deals, or any that decided or blocked) and the thresholds are assumed. The value case is Unlocks: it names what the catalogue products need from this map, and what else some of them need; it is not a forecast of sales. The measure-later share is counted from the source named on Inputs (CRM: contacts linked to each opportunity, with meetings and site visits logged against the opportunity by date), 40 open deals now, with a two-way spot-check; the realistic change (30 points), the extra meetings per deal and the owner's target win rate are assumed, so any £ figure is an illustration.

The deal and role records identify people at customers, including judgements such as who blocked a deal: they are kept under a legitimate interests assessment, business contacts and interviewees are told in a privacy notice, and they are kept for 36 months. Not legal advice. The map is finished when it is handed over. Add each closed deal as it closes and review the cards once a quarter.

The months after

What the agent logged and flagged from November to February, and what people decided

Fictional, continuing the example: Wrenstall Automation, its deals and every figure here are invented. Deals are shown by reference and people by role. The counts were tallied from the story, not by rerunning the workbook; seven of the agent’s log rows are on the Agent sheet of the worked-example workbook.

Wrenstall chose done for you. At handover Geoff set up the buying group agent in Wrenstall’s own account, in an AI assistant the business already uses, with its IT’s agreement: connected read-only to the CRM’s opportunities, contacts and logged meetings, the quotation log and the sales team’s diaries, and writing only to the workbook and one shared drafts document. Before it ran, Wrenstall confirmed its legitimate interests assessment and privacy notice covered the agent and recorded a data protection impact assessment, because the deal records judge identifiable people at customers deal after deal (not legal advice). It started on go-live, Monday 2 November. The two agent settings are on Inputs: a deal flagged 5 working days before its proposal is due if the gate isn’t met, and every closed deal’s roles confirmed by its salesperson within 10 working days. The run details are on the Agent sheet: the managing director reads the weekly note, the sales director approves any change to a card, and the office manager, who logs no meetings, runs the spot-check. The salespeople sent their own meeting requests and confirmed every deal the agent logged. Nothing left the business, and no decision changed, without one of its own people approving it. At the check-in on Wednesday 2 December, 30 days after go-live, Geoff read the log with the managing director on screen in Wrenstall’s account; his upkeep of the set-up ended there.

Deals logged
9
closed November to January; 8 confirmed by the salesperson within 10 working days
Gate flags
6
proposals due without the gate met; 4 met it before the proposal went
Thin cards
2 to 1
S1 Procurement now has 3 items of evidence
Early read
14 of 31
open deals with two or more roles met before the proposal (45%), against 8 of 40 (20%)
DateWhat the agent logged or flaggedConfirmed or decided byWhat happened
2 Nov 2026
Flag
WA-26-081 (S1, new robot cell; open)
Proposal due in 5 working days with only the operations manager met. Drafted a request for a payback review with the MD and the financial controller.
Sales directorSent his own version; met both on 6 November; proposal went a day late, on 10 November.
9 Nov 2026
Logged
WA-26-074 (S2, upgrade; lost 28 October)
Logged from the CRM and diaries; who decided and who blocked left Not known.
Sales engineer (ran the deal)Confirmed; procurement blocked on payment terms, from the customer’s loss email.
16 Nov 2026
Logged
Card: S1 Procurement
WA-26-078 logged: the card’s evidence rises from 2 items to 3, no longer thin.
Sales directorCard read against the new deal; no change.
30 Nov 2026
Flag
WA-26-079 (S1, controls; lost 13 November)
Roles not confirmed 10 working days after closing.
Sales directorConfirmed from the CRM and the loss email on 1 December.
7 Dec 2026
Flag
WA-26-083 (S2, new cell; open)
A contact’s name in the CRM notes kept out of the workbook: the row says “Finance”.
Sales directorSales team asked to note roles, not names, in the CRM too.
14 Dec 2026
Logged
WA-26-086 (S2, upgrade; open)
Gate met before the proposal (procurement and operations).
Sales directorProposal sent on 17 December.
1 Feb 2027
Logged
Measure later: early read
Open deals counted the same way as the baseline: 14 of 31 with two or more roles met before the proposal. Ten drawn at random for the spot-check.
Managing directorSpot-checked by the office manager; entered on the Value Case on 4 February.

The early read is an early look, not a verdict. The share rose from 20% to 45%, a change of 25 points, but with about 40 deals at each count the smallest change it can show is 28 points: this is inside what chance could do, so it is read again at the re-measure on 2 May 2027. The spot-check found 21 of 22 logged roles confirmed and 2 met but not logged, so the rise is not the logging alone. Recall fell as a share of the deals logged, because every new deal came from the CRM and diaries. Confirming a deal took a salesperson about 5 minutes, and reading the weekly note the managing director about 10 (estimates). The months after add nothing to the value case above.