Consulting, Sold Cold
The first client, the hardest category, and the engagement where the whole machine was learned.
- Timeline
- 16 months · 24 campaigns
- Role
- Built and ran the programme; solo at first, with an Outlyst team behind the second half
- Status
- Client engagement
- Primary stack
- Clay · Apollo · Smartlead · Google Maps
Headline metrics
- What it tackles
- A Sydney management consulting and custom software firm needed outbound in the single hardest category to sell by cold email: operational improvement, pitched cold to SME owners who believe they already know how to run their company.
- What it delivers
- A 16-month, 24-campaign programme across Australia and MENA: 38,062 prospects, 90,981 emails, 2,340 replies at 2.57% on emails sent, 116 companies formally interested, and a settled answer on which audience and which offer carry the category.
The problem in full
The firm sells operational improvement to business owners, and the custom software those changes run on. Management consulting is one of the hardest things in B2B to sell by cold email. The benefit is structural and long term, so nobody can put a number on better processes against this quarter's revenue. And a business owner's instinct when a stranger offers to improve how they run their company is that they already know how to run their company.
Two more obstacles sat on top of the category. The firm sells into 8 sectors across two service lines, and the operational strain looks different in each, so there was no single list to buy: a list built on one filter would have been wrong for most of the market. And the decision crosses three desks, felt by operations, procured through technology, signed by the owner, so a message that reaches only one of the three tends to stop there.
This was also the first client. The engagement began before any track record existed on either side, opened in spreadsheets with days of manual research per list, and had to earn every system it later ran on. It is the only engagement that spans the whole arc from manual research to AI workflows, which makes it the honest origin story of everything else on this site.
System design
The programme ran 24 campaigns in five phases over 16 months, and each phase changed one variable while holding the rest, stepping outward from the firm's core market into new segments and new geographies until the programme reached full scale. The first attempt was solo; an Outlyst team sat behind the second half, which is where the volume came from. What began as 5 days of manual spreadsheet research per list closed as AI-personalized campaigns, and the distance between those two endpoints is the machine every later engagement inherited.
Underneath the phases, four workflows stayed fixed. Qualification ran before research, never after, because research is the expensive step: AI agents read each company's website and public profile for the signals that predict need, manual process, fragmented systems, compliance gaps, scaling strain, and anything that failed was dropped before a research credit was spent on it. Contact prioritisation then built a ranked ladder of two to three decision makers per company, operations, technology, ownership, so the most relevant person heard from the campaign without the company receiving four copies of it.
The copy had one bar: a reader should be able to tell somebody actually looked at their business before writing. Every email opened on one specific, checkable fact about the company or the person, and every campaign, all 24, no exceptions, offered a complimentary assessment of the prospect's own operation before anything was asked in return. In a category where the value is abstract, removing all cost from the first step was the only reliable opener found. A postscript invited a plain no, because in cold email a clear no beats silence, in the pipeline and in the sending reputation.
Delivery and capture were built like infrastructure. Separate warmed sending domains kept away from the firm's own, several mailboxes per domain, SPF, DKIM and DMARC on all of them, rotation wave by wave, and every sequence stopping the moment a human replied. Replies came back through webhooks into one correspondence tracker holding the prospect, the campaign source, the reply category and the full thread, read and tagged by hand. That tracker is the source of every positive number in this study.
The calls that made it work
A free assessment on every campaign, no exceptions
Better processes is an abstraction a business owner cannot price, so the first email created the comparison instead of assuming it: a complimentary assessment of the prospect's own operation, given before anything was asked for. Across 24 campaigns it was the one constant, and it is the only opener I found that reliably works in a category this abstract.
When the offer plateaus, change the offer, not the subject line
The consulting angle plateaued, and the answer was not a better consulting email. The firm also builds custom software, so the campaigns moved to a concrete pitch: you are paying every month for several separate subscriptions that do not talk to each other, and that could be one system you own. A stack of software bills is something the reader can check against a bank statement. Same firm, same buyers, different door.
Qualify before you research, ladder before you send
Paying to research a company you then discard is the one mistake that scales badly, so AI qualification ran first and everything that failed it cost nothing further. Inside each qualified company, a ranked ladder capped contact at two or three people, because flooding a company is the fastest way to turn it into a former prospect. The result is that 38,062 prospects produced 2,340 replies without burning the market the firm still sells into.
Trade rate for reach, deliberately and in the open
Narrow Australian lists early in the programme ran at around 5% on small volume. Widening into MENA and into five-figure sends brought the rate to around 2%, and that trade bought the number the engagement is judged on: the widest phases returned 109 of the 116 interested companies. The best campaign at real scale did both, Campaign 21: 4,724 prospects, 545 replies, 32 interested companies from one send.
The results in full
Across 24 campaigns and 16 months, 90,981 emails reached 38,062 unique prospects and returned 2,340 replies, a 2.57% reply rate counted the strict way, replies over emails sent, this engagement's own convention. 116 companies formally expressed interest in management consulting bought cold, the number the programme is judged on, and the correspondence tracker holds 217 positive records behind them, every logged reply except the refusals: 50 interested, 54 who requested a meeting, 62 requesting more information, 32 asking to be contacted later, 12 routed straight to the founder's desk and 7 looking into the market. Counted on stated interest alone the tally is 185. Among the companies that replied with interest: Ray White, Ecosa, Netafim and Chargeasap.
The firm kept more than the pipeline. It kept a settled targeting answer, proven across all 24 campaigns and kept private to the firm; a proven second offer, custom software, that reopens the same market whenever the consulting angle cools; and a warm layer of 85 prospects tagged not interested but potential lead, which in a category where deals take a year is an asset that did not exist before the campaigns ran. In the client's own account, the level of response this programme produced had been missing for years before it.
Continue reading
Outbound in Norwegian
A 25.26% reply rate across 18,062 Norwegian prospects, in a language I do not speak.
Rebuilding a Dry Market
The addressable market ran out, so I scraped 70,000 records and built a new one.
One System, Two Payment Rails
Two payments platforms sold to regulated finance across Europe on one shared system.