Outbound Prospecting Plan · October 2026

Brindlecote Precision Sheet Metal Ltd

Laser cutting, CNC folding and fabricated sheet-metal assemblies for OEMs, from a site in Redditch. 48 staff.

Worked example. Brindlecote Precision Sheet Metal and its figures are fictional. All figures come from the worked-example spreadsheet: ALL CHECKS PASSED (with 8 notes).
The headline

About 4.2 new OEM customers a year, not the 12 wanted, and not paid back within three years

Year-one net
−£26,944
−£29,740 if it works half as well; year two £3,451, year three £21,590
New customers
4.2
a year after the list and hours limits (6.6 as planned); 2.1 in year one, the first in month 6
Cost per new customer
£6,245
as planned, £674 per meeting, selling time and cash
Cash
£5,520
a year for data, tools and compliance, plus £3,173 once; selling time £27,104 a year

Verdict: As planned, 450 approaches a month (350 fresh from the list) should bring about 5.1 meetings a month and 6.6 new customers a year. The list and the hours sustain 86% of that, and second approaches reply less, so expect about 4.2 a year. Each is worth £8,960 of gross profit over its first-year orders and £7,500 counted later, against £5,520 of cash and £27,104 of selling time a year. Year one: 2.1 new customers, the first in month 6, a net of -£26,944 (-£29,740 if it works half as well); year two £3,451; year three £21,590. The running total is not positive within three years: outbound at this scale is an investment over several years, not a quick return. The target of 12 a year needs about 641 fresh contacts and 119 selling hours a month, against 300 contacts a month the list allows and 77 hours available: not reachable as planned. Best for the money: “Cold email: four-step sequence to engineering and purchasing managers (with its follow-ups)”. Fix the 4 legal checks on Compliance before anything is sent.

Channel by channel

As planned, at full activity

ChannelMeetings a monthNew customers a yearCost a monthPer new customerPayback per new customerGross profit per £, with follow-ups
Cold email: four-step sequence to engineering and purchasing managers
Email · 250 a month · 2.0% reply, 45% to a meeting own or industry rates
1.92.5£1,230£5,95513 months3.05
LinkedIn: connection request then two messages
LinkedIn · 80 a month · 2.5% reply, 50% to a meeting assumed rate
0.81.1£639£6,96014 months2.37
Phone: follow-up call to people in the email sequence at the best-fit OEMs
Phone · 100 a month, follows up row 1 · 12.0% reply, 20% to a meeting assumed rate
2.02.6£1,077£4,89012 monthswith row 1
Letter with a cut-and-folded sample to the top OEMs
Letter · 20 a month · 3.0% reply, 50% to a meeting assumed rate
0.30.3£465£16,873Never0.98

The follow-up call only works because of the email sequence, so the two are judged together: as a pair they are the best value. Payback per new customer counts from the first approach: a new OEM’s first-year orders take a year to arrive, so even the best channel takes about a year. The sample letter costs more than it brings, and its reply rate (3%) is assumed above the cautious 1%: send it to twenty accounts and measure before deciding.

When it pays

The running total turns positive not within three years

369121518212427303336

Running total by month over three years (£): lowest −£28,904, after 36 months −£1,903.

Month one carries the set-up (sequences, scripts, the sending domain), the compliance fixes, list cleaning and legal review; months one to three build up the mailboxes and the routine; meetings follow a month after the first approaches, first orders four months after that, and each new customer’s first-year orders arrive over twelve months, with the later work counted over years two and three. Year one carries the build-up and the one-off costs; the plan only pays back as the customers won in year one keep ordering.

Against the target

641 fresh contacts a month needed for 12 new customers a year

No: 6,536 valid fresh contacts a year needed against 3,060 the list allows; 119 selling hours a month against 77 available. Second approaches would bring less.

The plan already uses 350 fresh contacts a month; the list of 1,800 decision-makers, 85% valid, allows 3,060 valid approaches a year if each is approached again after six months, and half of those would be second approaches at half the rate. Twelve new customers a year needs a bigger list, better rates measured from the first months, or a lower target. Rates should not be raised to make it fit.

Compliance

4 Law to fix before sending; 0 blocking (no fix planned); 3 good practice to fix

CheckKindStatusFix
Unsolicited emails and LinkedIn messages go only to corporate subscribers (companies, LLPs, public bodies); sole traders and partnerships only with their consent (PECR: LinkedIn messages are likely to count as electronic mail)LawPass–
Every message names the business, with a company's registered name, number and office in emails (Companies Act trading disclosures), and gives a simple way to opt out; opt-outs are honoured across every channel within days (PECR, UK GDPR)LawPass–
A legitimate interests assessment for prospecting is written and kept (UK GDPR)LawFail3 h
The source of every contact's data is lawful and recorded, and privacy information is given at the first approach (UK GDPR)LawNot known2 h
Phone numbers screened against the TPS and CTPS before every calling round (PECR)LawFail1 h
Email open tracking (tracking pixels) switched off, or used only with consent (PECR regulation 6)LawFail0.5 h
List de-duplicated, with existing customers and open deals taken off and the opt-out list applied before each roundGood practiceFail2 h
Cold email sent from a separate domain (never the main company domain) with SPF, DKIM and DMARC setGood practicePass–
Sending domain checked against the main blocklists before sending and monthlyGood practicePass–
Bounce rate under 2% on a test send to the listGood practiceNot known1 h
No more than the daily limit of emails per mailbox while warming upGood practicePass–
Spam complaints monitored and kept under 0.3% (the level Gmail's sender guidelines set)Good practiceNot known1 h
LinkedIn's terms on automation respected (no scraping, no unapproved automation tools)Good practicePass–

Nothing is sent until the legitimate interests assessment is written, the source of every contact is recorded, the call list is screened against the TPS and CTPS, and open tracking is switched off in the sending tool. The list is de-duplicated with existing customers and open deals taken off, the email list is verified before sending (£300), spam complaints are watched in the mailbox providers’ tools, and the documents are reviewed by a solicitor (£600). A checklist, not legal advice.

How confident are we?

What this rests on

Each channel’s contacts are reduced to the 85% with valid data, then taken through interested replies, meetings, proposals and wins at the rates on Channels: 62% of the new customers rest on at least one assumed rate, the rest on the business’s own results or an industry figure. 10% of wins are taken off as likely anyway (assumed). Each new customer’s first-year gross profit arrives over 12 months from the first order, and 50% of £15,000 later over the next 2 years. Activity builds up over 3 months; meetings follow 1 month after the approach and orders 4 months after that. Selling time is costed at £45 an hour, with 40 set-up hours in month one; meetings are those held, no-shows taken off. The list allows each valid contact an approach every 6 months; second approaches are assumed to reply at 50% of the first rate, and the list to be renewed at its present size each year. Where the list or the hours can’t sustain the plan, every month is scaled down to what they can (86% here). Compliance is a checklist, not legal advice.

Run the first three months as a measurement: replace each assumed rate with Brindlecote’s own, then re-run the plan before committing to the full year.

The months after

What the agent flagged, and what people decided

Brindlecote fixed the four legal checks first and started the plan on 2 November 2026 (Inputs row 46). It chose the guided route, and the agent was set up in an AI assistant the business already uses, with read-only access to the sending tool, the CRM and the TPS and CTPS screening record. LinkedIn activity is logged in the CRM by the people who send it; the agent never touches LinkedIn. Every Monday at 8:00 it compares each channel's approaches, the bounce rate, complaints, opt-outs and meetings held with the month of the plan they fall in, and sends the sales director a short note. It uses the rules on Inputs: a channel is flagged below 80% of the plan two weeks running. Nothing left Brindlecote, and no decision changed, without one of its own people approving it. These events continue the fictional example.

DateWhat the agent foundWhat was decided
9 Nov 2026
No flag
The first week: 31 approaches against about 30 planned; bounce rate 1.2%.Nothing needed.
23 Nov 2026
Flag
Phone follow-ups at 3 and 2 a week, against about 7 planned, two weeks running.Calls moved into fixed slots in the estimator's diary; numbers screened before the round.
7 Dec 2026
Flag
A 3.4% bounce rate on list batch B, above the 2% limit.Batch B verified again and 19 addresses taken off before sending resumed.
14 Dec 2026
Flag
Someone who had opted out by email was still on the call list.Taken off and marked do not contact the same day; opt-outs now checked before each calling round.
4 Jan 2027
No flag
Own rates after two months: email 1.6% against 2.0% in the plan; LinkedIn 3.1% against the cautious 2.5%, from few requests.Too few contacts to replace any rate yet; look again in March.
1 Feb 2027
Flag
One meeting held in four weeks, against about 2.2 planned for month 3.Two replies wanted samples first; samples sent. Look again in March.

At the 30-day check-in on 2 December, Geoff and the sales director read the log together. No rule needed changing.

What the agent added. The method is unchanged: same plan, same rates, same checks. The agent made sure activity was compared with the plan every week, so the calls that weren't happening were caught in their second week, and the bounce and opt-out problems were stopped before they damaged the sending domain or broke the rules. By March the business will have three months of its own rates, measured the same way, to replace the assumed ones. It is too early to say anything about new customers: the plan expects the first in month 6. No £ figure is claimed for the months after; the figures above are unchanged.