Direct

The Outbound and Account-Based Playbook

What this build is for

Outbound reaches a short list of named firms before they start looking, so you're the firm they remember when they do.

An owner whose margins slipped again this quarter rarely searches for help. They notice it, worry about it, and talk to their accountant. If Kestrel is going to be known at that moment, someone at Kestrel has to go to them first. That's the job of this build, and it's the only build that chooses its audience by name.

Two findings pull against a cold pitch. Gartner surveyed 646 B2B buyers in August and September 2025, and 67% said they'd prefer to buy without a sales rep (Gartner). And John Dawes of the Ehrenberg-Bass Institute estimates that only about 5% of business buyers are in the market for a category at any one time (Ehrenberg-Bass). So most cold messages reach people who aren't buying, and who'd rather not be sold to when they are.

Both findings point the same way. Outbound works when it picks the 5% by their signals and sends them something worth keeping, rather than a request for a meeting. It fails when it sends a sales pitch to the other 95%.

This build is in the Direct category: a person, reaching out. "Account-based" means the same work done for one named firm across several builds at once: the letter, the ads that firm's people see, the round table its owner is invited to. Its rule is the same as every build's: a message exists because a named firm is likely to be in a moment. A list bought because it was cheap isn't a reason.

Account management, for firms that are already clients, isn't here. It belongs to Customers.

The footprint, and what it reads from the chain

Outbound leads 1 cell and 7 moments: the owner's trigger, once for each of Kestrel's seven problems. The owner's question is the same each time, "Margins are slipping. Is it the plant?", but what set it off differs:

It supports 3 more cells (9 moments):

• COO · trigger, for each of the seven problems. Advertising leads it. Outbound adds the COO as the second named person at each firm on the list.
• Investor's operating partner · trigger ("The value-creation plan needs margin within 18 months."). Partners and channels leads it. Outbound writes to the operating partners of funds that buy in Kestrel's sectors.
• Investor's operating partner · comparing ("Have they done this in portfolio companies like ours?"). The pitch leads it, inside a deal. Outbound's portfolio pack is often what's on the partner's desk when they compare.

The owner is why this build exists. Owners decide in owner-led firms, rarely search, don't read guides, and ask their peers. Events reaches them in a room; Partners reaches them through their accountant. Outbound is the only build that reaches them at their own desk, by name.

What it reads, and never rewrites:

Its slice of the Score, and its own audit

Fifteen moments touch this build, and fourteen score zero. At the trigger, nothing Kestrel does puts it in front of an owner or a COO: none of the searches or places the Score checked named Kestrel, so Found is 0 and the gate is shut. Quality is low too. Nothing Kestrel publishes answers "Is it the plant?" in the owner's terms (Answered 5), and the next step is a contact form (Bought 15). Believed and Preferred weren't measured for these moments, so they're left out of the average.

One limit to be clear about: Known is measured across the whole market. A letter to 200 firms out of 1,400 moves it slowly. So this build is judged first by what happens on its named list, and the Score shows the effect later, as those firms start to find Kestrel by other routes.

The audit. What Kestrel's outbound did last year:

Eleven letters produced more than 9,600 emails. The difference wasn't the channel. Each letter went to a named person, for a reason, from someone they knew.

Step 1: the shape, the account universe

The shape is every firm in your segments, counted, before anyone decides who to write to. Kestrel's segments come from its GTM strategy: manufacturers with 80 to 400 staff, in four sectors, in the region it can reach in a day.

The 210 investor-backed firms belong to about 60 funds. 25 of those funds own two or more manufacturers in these sectors, and their operating partners are a short list of their own. Group-owned firms decide at group level, often abroad, so they stay in the shape but rank lower.

Now apply the 95:5 rule. About 5% of 1,400 firms, around 70, are in the market at any one time. Nobody knows which 70. The job of the next step is to find signals that make a good guess.

The shape splits into three tiers, by how much work each firm gets:

Firms move up when a signal appears, and back down when 90 days pass without a reply. Forty is a ceiling, not a target. It's the number three partners can write to well, and follow up personally.

Step 2: coverage, one signal per moment

The coverage unit is a signal: something you can see from outside that suggests a named firm is in a moment now. For each moment in the footprint, ask what would show it, where you'd see it, and how much you'd trust it.

Three checks before a signal puts a firm in the one-to-one tier:

It's recent. Within the last 90 days, or 100 days after a deal. Older signals mean the firm has probably acted already.

It's specific. "Hiring" isn't a signal. "Hiring a second production planner, eight weeks after a delivery complaint in the trade press" is.

Someone could explain it to the owner. If a partner couldn't say in one sentence why they're writing, the signal isn't strong enough.

When Kestrel first ran this across the 1,400 firms, 52 showed a signal that passed all three checks: 11 deals, 9 new lines, 14 margin falls, 8 stock rises, 6 recalls or quality hires and 4 third-shift hires. That's close to the 70 the 95:5 rule predicts. The top 40, ranked by how strong and recent the signal was, became the one-to-one tier.

Step 3: gaps, biases and jobs

One pattern: every job starts with a reason the firm would recognise and a page it would keep. Nothing goes to a firm without a signal from the last 90 days, and nothing asks for a meeting. The letter's only request is a reply, if the reader wants one.

Step 4: reconcile, the outbound plan

Rejected, with reasons:

The bought list and the agency sequence. 9,600 emails, four meetings, no reviews, and 15 people who now associate Kestrel with spam. The money pays for a researcher's time to find signals.

Connection requests followed by a pitch. It reads as a trick, because it is one. Partners can connect after a conversation, not before.

Cold calls to owners. No moment asks for one, and owners screen their calls. A reply to a letter is a better start than an interrupted afternoon.

Gifts in the post. They raise the question "what do they want from me?" before the reader knows who Kestrel is. In firms with a procurement team, a gift can be a problem in itself.

Bought intent data. Kestrel's category is too small for third-party intent signals to be reliable, and the public signals above are free. Review it if the shape grows past a few thousand firms.

Sending to all 1,400. About 95% aren't in the market. The other builds reach them, at lower cost, without spending the firm's name.

The anatomy library

Four rules hold for every message:

A reason they'd recognise. Every letter names the signal it's responding to. If the reason would surprise or unsettle the owner ("we noticed your recall"), it isn't used.

Something to keep, nothing to ask. The enclosure is useful whether or not they ever speak to Kestrel. The only request is a reply.

Two letters, then stop. A third letter to someone who hasn't replied tells them Kestrel doesn't listen. The firm comes back only with a new signal.

Within the rules. The law on business emails and post differs by country. Write to a named person in their business role, say where you found their details, and make opting out one click or one line. Check your own country's rules before the first send, not after a complaint.

Briefs

One brief per account in the one-to-one tier, written before anything is sent. This is the brief behind the finished example below.

ACCOUNT Harrowfield Precision: metal and engineering, 190 staff,
owner-led
SIGNAL Bought Kellow Machining (60 staff) on 14 January.
               Source: trade press, 16 January; company register, 21 January
MOMENT Owner · trigger · after an acquisition
               "We need to show margin improvement within 18 months."
NAMED Richard Hale, managing director and owner
               Nadia Shah, operations director (second letter)
WARM ROUTE Checked: their accountant isn't one of Kestrel's partners;
               no past client works there. Cold, by letter.
ENCLOSE "The first 100 days on the floor" (one page)
CLIENT STORY A 150-person machining business, after buying a competitor
               (unnamed; figures confirmed by the client)
SIGNED BY Jo Brady, partner (covers metal and engineering)
FOLLOW-UP Three to four weeks later: the March owners' round table
IF THEY REPLY The pitch, within one working day: a call to look at one
               number together
STOP After the second letter, if no reply

And the brief for the operating partners' pack, which runs once a year:

AUDIENCE Operating partners at the 25 funds owning two or more
               manufacturers in Kestrel's four sectors
MOMENTS Operating partner · trigger: "The value-creation plan needs
               margin within 18 months."
               Operating partner · comparing: "Have they done this in
               portfolio companies like ours?"
WARM ROUTE Partners and channels first: 6 of the 25 know a Kestrel
partner
THE PACK Results across investor-backed clients (signed by
Credentials);
               the first six weeks; a sample of the board report
SIGNED BY The managing partner
WHEN Once a year, and within 30 days of each new manufacturing deal
MEASURE Introductions to portfolio companies within 6 months

The finished example: one account, written in full

The letter, its enclosure, the COO's letter, the follow-up, and the handover when the owner replied. Harrowfield Precision, Kellow Machining, the people and Kestrel are invented.

1. The letter, to Richard Hale, managing director and owner of Harrowfield Precision, by post and by email, 12 days after the deal.

• Dear Richard,
• Congratulations on bringing Kellow Machining into Harrowfield. I read about it in the trade press in January.
• I'm writing because the months after a purchase are when owners most often ask us for help, and it's nearly always about the same thing: the two plants' numbers don't add up the way the plan said they would. In the six engineering businesses we've worked with after an acquisition, first-year margin came in an average of four points below the plan, and most of the gap was on the floor, not in the overheads.
• I've enclosed one page we wrote for owners in your position: what usually slips in the first 100 days, and three things you can check this month without anyone's help. It's yours to keep, whether or not we ever speak.
• If any of it is useful and you'd like to compare notes, reply to this letter or email me at jo.brady@kestrel.example. If not, I'll write once more in a month or so, and then leave you in peace.
• Jo Brady, Partner, Kestrel Advisory

2. The enclosure: The first 100 days on the floor.

• The finding. After a purchase, the plan assumes both plants will run at the better plant's level. In the first year, they usually run nearer the weaker one's.
• Three things to check this month:
• 1. One number, both plants. Output per labour hour on your most similar product. If
• the gap is more than 15%, the plan's savings are probably in the wrong place.
• 2. Both floors at shift change, in the same week. Time how long it takes to make the
• first good part. That's capacity your accounts won't show.
• 3. One question to both planning teams: "What stops us shipping on time?" If the
• answers differ, the plants run on different rules, and a merged schedule inherits the worst of both.
• At a client. A 150-person machining business bought a competitor. Output per labour hour differed by 22% between the two plants. A year later the gap was 6%, and margin was three points up.
• Where this comes from. Kestrel's reviews in six engineering businesses after an acquisition, with each client's figures confirmed by the client.

3. The COO's letter, to Nadia Shah, operations director, the same week.

• Dear Nadia,
• After an acquisition, the operations director usually inherits the hardest part of the plan: two plants, one set of targets, and a board expecting both to move at once.
• I've enclosed the first page of our guide to running two plants as one. It's written for the person who runs the floor, not the board. I've written to Richard as well, with a shorter page for owners.
• If anything in it is useful, reply and I'll send the rest. If not, no need to reply.
• Jo Brady, Partner, Kestrel Advisory

4. The follow-up, to Richard, four weeks later. It brings something new, rather than "just checking in".

• Dear Richard,
• One more thing that may be useful, and then I'll leave you in peace.
• On 12 March the Metal and Engineering Association is hosting a dinner for 14 owners. Helen Dorsey, who runs a 230-person fabricator, will talk about when an owner needs outside help, and when not. Nothing will be sold. If you'd like a place, reply and I'll ask the association to send you an invitation.
• Either way, I wish you and the Kellow team well.
• Jo

5. Richard's reply, and the handover, the same day.

• Richard's reply: "Jo, I did your first check. 19% gap. Not what I was told when we bought it. Yes to the dinner. Can we talk before then?"

TO Mark Ellison, head of delivery (runs the pitch for this
account)
ACCOUNT Harrowfield Precision; Richard Hale, owner
WHAT HE SAID "I did your first check. 19% gap. Not what I was told when
               we bought it. Yes to the dinner. Can we talk before then?"
SIGNAL Bought Kellow Machining, 14 January
MOMENT Now: owner · options, "Do we need outside help at all?"
SENT SO FAR Letter and enclosure; COO letter (no reply from Nadia);
               round-table follow-up
NEXT STEP A 30-minute call to go through his 19% together. No proposal.
DINNER Place requested from the association

What to notice. The letter names the reason for writing in its first line, and the reason is good news, not a weakness. The enclosure is useful to Richard even if he never replies, and it gave him a check he could do himself; the reply came because the check found something. The follow-up offered a room full of other owners rather than a meeting with Kestrel. And the reply moved Richard from the trigger to the options moment, so the next step comes from the pitch, not from Outbound. One account; about 40 at a time.

Templates

The account list (one row per firm in the one-to-one tier)

The account brief

ACCOUNT [firm: sector, size, ownership]
SIGNAL [what, when, and where you saw it]
MOMENT [chooser · step · problem, and their likely question]
NAMED [the owner; the COO or operations director]
WARM ROUTE [partner, accountant, association, past client; or "cold"]
ENCLOSE [the one page, and why it fits this firm]
CLIENT STORY [a client like them, and whether they may be named]
SIGNED BY [the partner for this sector]
FOLLOW-UP [the new thing the second letter brings]
IF THEY REPLY [who in the pitch, and the smallest next step]
STOP [after the second letter, if no reply]

The signal log (one row per signal found, whether or not it's used)

The signal log keeps the tier accountable. It shows how many signals were found, how many passed the three checks, and how many became letters.

In the campaign plan

Most builds follow the campaign calendar. Outbound mostly follows the signals, which arrive when they arrive. What the Campaign Engine gives it is the enclosures: each campaign produces a page worth posting to an owner, and Outbound takes it to the named firms whose signal fits.

The account plan joins the builds up for each of the 40 firms. While Jo is writing to Richard, Advertising shows the "after an acquisition" page to people at Harrowfield, Events holds a place at the March dinner, and the Website's acquisition page is the one the letter links to. One firm sees one story from three directions in the same month.

Running it

Who owns what. At Kestrel, a researcher (two days a week) owns the signal log and the account list. Three partners, one per sector group, own the letters for their accounts and sign them personally. The marketing lead owns the enclosures, with Content, and the account plans, with Advertising and Events. The managing partner owns the operating partners' pack. Whoever runs the pitch owns every reply from the day it arrives.

Partners make the letters their own. The researcher drafts the brief and the marketing lead drafts the letter. The partner rewrites it until it sounds like them, then signs it. An owner can tell a letter a partner wrote from one they were handed to sign.

One opt-out list for every build. Anyone who says no to Outbound is also off Email's lists and Advertising's account audiences. Automation keeps the list, so nobody has to remember.

Common mistakes

The bias to watch in the team: base-rate neglect. About 95% of firms aren't in the market at any one time. Teams forget it in both directions: they read a big list as a big opportunity, and a quiet month as a bad letter. The base rate says to send fewer letters, to firms with a reason, and to judge them over a year.

The business check and the measures

Revenue: owner-led firms are 830 of the 1,400 in the shape, and their owners are the chooser least likely to find Kestrel on their own. A review won from the named list pays for a year of the researcher's time.

One truth: every number in a letter or an enclosure is one Credentials can sign, and every client story is confirmed by the client. Letters never guess at a firm's problems from outside.

The proof is real: the enclosures are built from Kestrel's own reviews. If a sector has no confirmed result yet, its letters use the benchmark instead of a client story.

Capacity: about 10 new accounts a month means around 30 letters, including COO letters and follow-ups. At about 45 minutes each, with the partner's rewrite, that's roughly 22 hours a month across three partners, plus the researcher's two days a week. That's less than the agency and the list cost last year.

The Loop for this build

Each quarter, the Loop (step 11) asks: did the firms we wrote to move from the trigger to a conversation, and which signals told us the truth?

What Kestrel's first year would look for. Replies from at least one in ten of the one-to-one accounts. At least four reviews won from the named list, against one last year. A third of the named firms turning up later by another route. And no complaints: the clearest sign that an owner found a letter welcome rather than intrusive.

Shapes by business type

Feeling a bit overwhelmed?

You should be, a little. One firm: 1,400 firms counted, a researcher checking signals every week, three tests for each, 40 account briefs, a letter and an enclosure for each, a second letter to the COO, a follow-up that brings something new, a handover within a working day, and a close-out when it's time to stop. Most firms buy a list, load a sequence, and count the emails.

RocMode does part of the heavy lifting:

The people you'd write to, and their moments. Your choosers and their moments, mapped from what's known about your brand, so every account brief starts from who's likely to be asking what, and why.

The segments your account list starts from. RocMode's GTM strategy sets out who you sell to, so the shape is counted from your segments, not from whatever list is cheapest.

What happens when someone replies. RocMode writes a demo script and an objection sheet for each buyer, and every answer cites its proof or names the proof still to collect.

Join the RocMode beta