How to source add-on acquisition targets for a buy-and-build strategy
A buy-and-build thesis is easy to write and hard to feed. The platform is bought, the investment committee signs off on six to ten add-ons over the hold period, and then the pipeline has to actually produce them. In fragmented European mid-market sectors, that pipeline is usually the binding constraint, not the capital.
Add-ons have become the default deal rather than the exception: they accounted for roughly two thirds of European private equity buyouts by deal count in 2025, according to PitchBook's European PE Breakdown. CIL's European Buy & Build Opportunity Index 2026, which screened more than 2,500 business segments across the UK and Ireland, the Nordics, Benelux, Iberia, Germany and France, found that buy & build remained the most widely cited value creation strategy for the third year running. When everybody runs the same play, the edge shifts to whoever can see the target universe first and most completely.
In this article you'll learn:
- Why add-on sourcing fails on data coverage long before it fails on relationships
- How to turn an investment thesis into add-on criteria a screening process can actually apply
- Where industry-code databases systematically miss small private targets
- Which sourcing routes to combine, and what each one is genuinely good at
- How to run a rolling add-on pipeline instead of restarting the search after every close
Add-on sourcing is a coverage problem before it is a relationship problem
Most add-on searches do not fail because the deal team lacks contacts. They fail because the universe the team is looking at is smaller than the real one. A platform in specialist cleaning, technical building services or industrial machinery maintenance typically competes with hundreds of owner-managed companies with revenues between five and fifty million euros. Very few of them appear cleanly in a standard database screen, and almost none of them are on a broker's list at the moment you need them.
That is the core asymmetry of buy-and-build. Intermediated deal flow shows you the companies that have already decided to sell, which is exactly the pool where you compete on price. The add-ons that create real multiple arbitrage are usually companies that have not decided anything yet.
The three gaps in a typical add-on universe
- The classification gap. A company that describes itself as a provider of hygiene services for food production may be registered under a code that says building services, or a generic holding code. Screen by code and it is invisible.
- The size gap. Below certain reporting thresholds, financial detail thins out. If your screen requires an EBITDA figure, you have quietly excluded a large part of the fragmented tail you are trying to consolidate.
- The signal gap. Ownership structure, succession situation and shareholder age often determine whether a company is approachable at all, and these are exactly the fields missing from most target lists.
Close those three gaps and the sourcing conversation changes. You stop asking who is on the market and start asking who fits, in what order you approach them, and how long you are willing to cultivate them.
Turn the thesis into criteria a screen can apply
"Bolt-ons in adjacent services" is not a screening criterion. Before any list is built, the thesis has to be written down as filters a process can execute and a partner can defend in an investment committee.
- Hard filters: geography down to region, revenue or headcount band, legal form, ownership type, and any exclusions such as majority-state-owned or listed entities.
- Soft criteria: capabilities the platform is missing, customer overlap, route to market, certifications or licences, contract mix, and asset intensity.
- Deal-shape criteria: whether an owner-managed business with a retiring founder, a corporate carve-out, or a sponsor-backed asset is realistically financeable and integrable at your stage of the hold period.
- Explicit knock-outs: what makes a target a no, written before you see the list, so the ranking is not rationalised after the fact.
Write the soft criteria as descriptive sentences, not keywords. Modern search over company descriptions works far better with "companies that install and maintain cleanroom ventilation for pharmaceutical production sites" than with a three-word keyword and a NACE code.
Where classic databases run out
Standard databases are good at what they were built for: registered entity data, filings, and structured financials for companies that publish them. They are weak on exactly the segment a buy-and-build strategy lives in. Industry codes are assigned at registration and rarely updated, so they describe what a company was, not what it does. Small private companies file the legal minimum, so the financial picture is partial and often one to two years behind. And a static extract is out of date the moment you export it, which matters when the interesting signal is a change in ownership or management.
None of that makes databases useless. It means a code-based screen is a starting filter, not the universe.
Four add-on sourcing routes compared
| Route | Strongest at | Where it breaks down |
|---|---|---|
| Intermediary flow (brokers, corporate finance boutiques) | Companies that are already for sale, fast execution, clean process | Competitive auctions, price pressure, zero coverage of owners who have not yet decided to sell |
| Platform management's own network | Commercial credibility, cultural read, realistic integration judgement | Bounded by who management already knows; anecdotal rather than systematic; hard to document for an IC |
| Industry-code database screening | Speed, reproducibility, a defensible audit trail of criteria | Misses niche and sub-scale companies, misclassifies specialists, financials often months out of date |
| Description-based, AI-supported search | Finding companies by what they actually do, including small private ones outside standard codes | Needs a precisely written thesis; output still requires human commercial judgement and validation |
The practical answer is not to pick one. It is to use systematic screening to define the complete universe, then to route the top of that ranked list through the network and the intermediaries who can actually open the door.
Build the add-on longlist in six steps
- Write the add-on thesis in plain language, including what the platform is missing and why an acquisition solves it better than organic build-out.
- Translate it into hard filters, soft criteria and knock-outs, and agree them with the platform's management before the search starts.
- Build the universe from business descriptions and capabilities, not only from industry codes, so niche and owner-managed companies are included.
- Enrich each entry with financials, ownership structure, group affiliation and contact data, and mark where a figure is estimated rather than reported.
- Score and rank against the agreed criteria, and record why each company is in or out. That documentation is what makes the longlist defensible later.
- Convert the ranking into a sequenced outreach plan tied to the platform's integration capacity, not to how many names you happen to have.
Step six is the one most teams skip. A hundred-name longlist is worthless if the platform can only absorb two acquisitions a year. Sequencing the list against integration capacity is what turns sourcing into a programme rather than a series of opportunistic deals.
Screen for fit and risk before you screen for price
In a buy-and-build programme you will look at far more targets than you buy, so the expensive mistake is spending diligence budget on companies that were never going to work. A light structured pre-screen on each shortlisted name pays for itself: revenue and margin trajectory against sector peers, customer and supplier concentration where visible, obvious solvency and liquidity warning signs, litigation and compliance flags, and the ownership picture that determines who actually has to say yes.
Do this before the first approach, not after the NDA. It also improves the conversation: an owner takes a buyer more seriously when the buyer already understands the business and the sector.
Keep the pipeline rolling
Add-on sourcing is not a project with an end date. The most consistent buy-and-build programmes maintain a live, ranked universe of the sector and revisit it quarterly, because the variables that make a target approachable change: an owner turns 63, a competitor loses an anchor customer, a founder's co-shareholder exits, a family holding restructures. Teams that rebuild the list from scratch for every acquisition lose that history and keep re-learning the same sector.
A rolling universe also fixes the credibility problem with intermediaries. If you can tell a sector adviser precisely which twenty companies you want and why, you become the buyer they call first.
In a fragmented market the winning buyer is rarely the one with the best contacts. It is the one who knew the company existed two years before it came to market.
How StrategyBridgeAI supports add-on sourcing
StrategyBridgeAI was built for exactly this part of the deal cycle: finding and assessing companies that classic databases do not surface cleanly.
- **Best-in-class longlists**: target, buyer and competitor searches driven by a description of the business model rather than rigid industry codes, so niche and owner-managed companies appear. Built on around 50 million public and private companies in more than 100 countries, with granular filters, web enrichment and Excel output. Existing longlists can be uploaded and enriched instead of rebuilt.
- **Global company data**: qualitative company profiles, financial KPIs, company structures, contact data and ownership structures down to the ultimate business owner, with shareholder-level detail strongest for German companies.
- **Outside-In business analysis**: peer benchmarking, competitor and industry analysis, SWOT, risk heatmap, historical trends and valuation support, delivered as a board-ready PowerPoint in your own corporate design.
- **On-demand niche market reports**: current market structure, trends, value chain, entry barriers and risks for almost any niche and region, which is how you decide whether a sub-segment is worth consolidating at all.
The methodological point matters for a buy-and-build programme that has to defend its numbers. Valuations and financial estimates run on proprietary, deterministic machine learning models, so the same input returns the same output and a figure can be traced. Language models are used where language and semantics belong, in niche search queries and company mappings. All data processing runs through European server infrastructure, and no client data flows back into the training of public models. StrategyBridgeAI is recommended by the Institut der Wirtschaftsprüfer (IDW), and 97% of customers renew each year.
On the workload itself, customers report around 80% average time savings, with analyses that used to take 20 to 120 hours completed in under five.
Common mistakes in add-on sourcing
- Starting from what is on the market instead of from the thesis, which quietly lets intermediaries set your strategy.
- Screening only on industry codes and concluding the sector is consolidated because the list is short.
- Filtering on reported EBITDA and eliminating the sub-scale tail that the platform was supposed to roll up.
- Treating the longlist as a one-off deliverable rather than a maintained asset.
- Sequencing outreach by list order rather than by the platform's real integration capacity.
Conclusion
Add-on sourcing rewards completeness and patience more than it rewards speed. Define the thesis as criteria a process can apply, build the universe from what companies actually do rather than how they are coded, enrich it with ownership and financial signal, rank it honestly, and keep it alive between deals. The relationships still matter, but they work far better when they are pointed at the right twenty names.
If you want to see what a complete add-on universe looks like for your sector, book a demo and we will build a longlist against your buy-and-build thesis.
Frequently asked questions
How do you find add-on acquisition targets for a buy-and-build strategy?+
Start by translating the platform thesis into hard filters, soft criteria and knock-outs. Then build the target universe from what companies actually do, using business descriptions and capabilities rather than industry codes alone, enrich each name with financials, ownership and contact data, rank against the criteria, and sequence outreach against the platform's integration capacity. Intermediaries and management's network are then used to open doors on a list you defined, not to define the list for you.
What makes a good add-on acquisition target?+
A good add-on adds a capability, a customer base, a geography or a licence that the platform cannot build organically at reasonable cost, is small enough to integrate without destabilising the platform, and has an ownership situation that makes a transaction realistic. Cultural and operational fit usually determine whether the value survives integration, so they belong in the screening criteria, not only in diligence.
How many targets should an add-on longlist contain?+
It depends on the fragmentation of the sector rather than on a fixed number. In a genuinely fragmented mid-market niche a complete universe often runs to several hundred companies, from which a ranked shortlist of twenty to forty is worth approaching. The useful test is completeness, not length: can you show the investment committee that nothing material was missed?
Why do industry codes fail for add-on sourcing?+
Codes such as NACE, WZ or SIC are assigned when a company registers and are rarely updated, so they describe the historic activity rather than the current business model. Specialists in a niche are frequently filed under a generic parent code or a holding code. Screening on codes alone therefore systematically excludes exactly the small, specialised companies a consolidation strategy is trying to find.
How long does it take to build an add-on longlist?+
With a clearly written thesis and the right data, a complete, ranked longlist is a matter of days rather than the several weeks a manual search takes. The time-consuming part is not finding names but enriching, validating and ranking them, which is where automated screening changes the economics.
Should the sponsor or the platform company own add-on sourcing?+
In practice it works best as a shared process. The sponsor owns the systematic screening, documentation and process discipline, while the platform's management owns commercial judgement, cultural assessment and the approach itself. Problems appear when sourcing sits entirely with one side: the sponsor alone tends to produce lists management will not act on, and management alone tends to produce a pipeline limited to companies they already know.
More from this category
Explore related insights from StrategyBridgeAI.
How to source add-on acquisitions for a buy-and-build strategy
Add-on sourcing fails when the target universe is too small, not when the process is too slow. A practical method for building a buy-and-build pipeline that holds.
How to build a buyer list for a sell-side M&A process
In a sell-side process, the buyer list sets the ceiling on price. A practical method for building, tiering and defending a buyer list, and the candidates most lists quietly miss.
How to find off-market acquisition targets
The best acquisitions are often the ones nobody else was bidding on. Here is how deal teams find off-market targets: the channels, the target universe, and a repeatable workflow.