Buy-and-build is the dominant value creation strategy in private equity. Across mid-market PE, more than 70% of platform companies execute at least one add-on acquisition during the hold period. The logic is sound: bolt-ons expand geographic reach, add product capabilities, consolidate fragmented markets, and drive multiple expansion through scale. The problem is not the strategy. The problem is sourcing the targets.

Most PE firms approach add-on sourcing the same way they approach platform sourcing – through broker relationships, inbound referrals, and the occasional conference conversation. This works when the add-on criteria are broad and the market is deep. It breaks down when the thesis is specific, the targets are owner-operated businesses that are not formally for sale, and the timing needs to align with both the platform's integration capacity and the target owner's willingness to transact.

Why add-on sourcing is different

Add-on acquisitions have a structural characteristic that makes them both more valuable and more difficult to source than platform deals. The target universe is defined by the platform's specific needs – geographic adjacencies, product overlaps, customer segment extensions, capability gaps. This means the universe is bounded and identifiable, but the targets within it are typically smaller, owner-operated businesses that are not on any broker's radar.

A $12 million revenue HVAC business in the mid-Atlantic is not going to hire Goldman Sachs to run a process. The owner is more likely to sell to someone who shows up at the right time with a credible offer and a clear explanation of why the combination makes sense. Finding that owner, reaching them directly, and starting a conversation before a broker gets involved – that is the add-on sourcing challenge, and it requires a fundamentally different approach than waiting for broker flow.

Step 1: Define adjacency criteria

The first step is precision in defining what "adjacent" means for the specific platform. This is where most internal BD efforts go wrong – the criteria are too broad. "Companies in our industry with $5M to $30M in revenue" is not a useful targeting framework. It produces a list of thousands with no way to prioritise.

Adjacency needs to be defined along at least four dimensions. Geographic overlap: which markets does the platform want to enter, and what radius constitutes a reasonable integration? Product adjacency: which services or capabilities would the platform add, and are there specific product lines that command premium pricing? Customer overlap: are there target companies serving the same end customer in a complementary capacity? Capability gaps: does the platform need a specific certification, technology, or talent base that acquisition would provide faster than organic development?

When these criteria are well defined, the target universe narrows dramatically. Instead of 3,000 vaguely relevant businesses, you end up with 300 to 500 companies that genuinely fit – each one assessable against a clear framework.

Step 2: Build the target universe

With criteria defined, the next step is mapping the actual companies. This is not a database pull. It is a research process that involves identifying businesses that meet the adjacency criteria, verifying ownership structures (is this founder-owned? family-owned? has it taken outside capital?), confirming that the business is the right scale, and identifying the decision-maker – by name, with verified contact information.

In a typical add-on campaign, we map 340 or more targets, verify ownership and decision-maker contact information for each, and score every company against the adjacency criteria on a three-tier fit scale.

This mapping process takes time, but it produces something that no database subscription can provide: a proprietary, verified, scored universe of acquisition targets that is specific to your platform's thesis. It is an asset that appreciates over time as data accumulates and intelligence compounds.

Step 3: The signal layer

A mapped universe is necessary but not sufficient. The question is not just "who fits?" but "who fits and is showing signs of timing?" This is where the signal layer transforms a static list into a dynamic pipeline.

For add-on targets, the most predictive signals include succession indicators – the founder is approaching retirement age, has not groomed a successor, and the business has no clear continuity plan. Growth stalls are another strong signal: revenue has plateaued after years of consistent growth, which often prompts an owner to consider strategic options. Competitive pressure matters too – when a direct competitor gets acquired by a well-capitalised buyer, remaining independents often reassess their position. And there are financial signals: a business with an upcoming loan maturity or an expiring lease on a key facility is facing a natural decision point.

Each signal on its own is suggestive. Two or three signals converging on the same target is a strong indicator that outreach will be well-timed.

Step 4: Sequenced, contextual outreach

The outreach itself is where the entire process either converts or collapses. A generic message to a business owner – even one who is genuinely considering a transaction – will be ignored or deleted. The owner receives dozens of these messages monthly. They all sound the same, and none of them demonstrate any actual understanding of the business.

Signal-informed outreach is different. The message references the specific context: "We've been following the growth of your Southeast operation and think there's a natural fit with [platform name], particularly given the complementary customer base in commercial construction." This is not flattery. It is evidence that the sender has done the work to understand the business, and it immediately differentiates the conversation from every other cold email in the owner's inbox.

The outreach is sequenced across channels – email, phone, and in some cases direct mail – with each touchpoint adding context rather than simply repeating the initial message. The sequence is white-labelled to the PE firm or the platform company, depending on the client's preference, so the owner experiences it as a direct relationship rather than a brokered introduction.

The results

In a recent add-on campaign for a mid-market industrial services platform, the numbers looked like this: 342 targets mapped, 312 owners with verified contact information, outreach sequenced over eight weeks. The campaign produced a 5.8% reply rate – compared to the 1% to 2% that generic outreach typically generates. Of those replies, 14 progressed to substantive conversations about a potential transaction. Three entered formal evaluation within the first quarter.

Those numbers reflect a single wave. The compounding effect is what makes the system valuable over time. Wave one produced 14 interested conversations, but it also produced data on which signal types drove the highest engagement, which messaging angles resonated, and which segments of the target universe were most responsive. Wave two, launched three months later with refined criteria, produced a 7.1% reply rate and 19 interested conversations from a smaller outreach universe.

The compounding effect

This is the structural advantage of systematic add-on sourcing. Each campaign is not an isolated event. It is an iteration on a growing dataset. The target universe is continuously maintained – companies that transacted are removed, new entrants are added, signal data is updated. The outreach sequences are refined based on response data. The signal model is adjusted based on which indicators actually predicted engagement versus which ones did not.

By the third or fourth wave, the system is operating at a level of precision that no ad hoc effort can match. The PE firm has a proprietary, continuously improving origination engine that is producing conversations with verified, pre-qualified add-on targets – conversations that their competitors are not having because they are still waiting for a broker to call.