The Synergy Number Nobody Checks
Every acquisition model in industrials carries a purchasing synergy line. In a buy-and-build platform it is often the single largest contributor to the value case. And in most deals it is produced the same way: take the combined direct material spend of the two entities, apply a percentage that has worked before, and put the result into the model.
That percentage is an assumption. It gets treated as an analysis because it sits in a cell next to numbers that were analysed.
Commercial, financial, legal and tax diligence each get a dedicated workstream with named owners and a data room. Supply chain gets a questionnaire and a management estimate, usually in the final fortnight, after the price has been agreed. The result is that the line item most likely to determine whether the deal works is the line item nobody stress tested.
What a real synergy analysis looks like
A recent engagement makes the shape of the work concrete. Two of the largest auto parts groups in the world, one European and one North American, were evaluating a combination. The question was how much purchasing synergy was genuinely available and from which categories.
The analysis ran four to six weeks. It covered 11 categories and 121 sub-categories, and identified $56M from the top 10 suppliers alone.
The number matters less than how it was produced. A synergy figure that cannot be defended line by line does not survive the first management challenge, let alone a lender.
Step one: match the SKUs
This is where synergy studies live or die, and it is the step most often skipped.
The premise of purchasing synergy is that two entities buy the same thing at different prices, and the combined entity can pay the lower one. Testing that requires knowing which items are actually the same item.
Branded parts make this easy. OE numbers provide a common identifier across both purchasing bases, and matching is largely mechanical.
Private label does not match on anything. There is no shared part number, no shared nomenclature, and frequently no shared supplier. The match has to be constructed from technical specification, using client documentation alongside publicly catalogued data, attribute by attribute, until two items can be declared equivalent with evidence behind the claim.
Skip this and the analysis compares two different parts and reports the difference as a saving. That is how a synergy number becomes indefensible.
Step two: normalise to net-net
List price is meaningless across two purchasing bases.
One entity buys ex-works from a supplier three hundred kilometres away with a volume rebate settled annually. The other buys delivered, in a different currency, with a rebate structure tied to payment terms rather than volume. Comparing invoice prices compares logistics and commercial structure, not procurement performance.
Everything has to be stripped back to net-net after freight, rebates and currency before any comparison is made. The normalisation work is unglamorous and it is the difference between a credible number and a coincidence.
Step three: extrapolate honestly
Matched spend is only ever a share of total spend. In a large industrial pairing, the matched base might be a fraction of the combined purchasing volume, concentrated in the categories where identification was easiest.
The identified saving therefore has to extend to the unmatched base, and how that extension is done determines whether the number is trusted. A combo-effect basis, where the observed price differential in matched items is projected across comparable unmatched spend, is defensible. A flat percentage applied to everything is not.
The assumptions behind the extrapolation should be validated by both management teams rather than asserted by the advisor. A synergy number the two operating teams have already argued with is a number that survives integration planning.
Step four: rank by difficulty, not by size
A synergy report sorted by value is a report that will disappoint.
A five million dollar gap on a dual-sourced commodity with two qualified suppliers is worth more than a fifteen million dollar gap on a customer-nominated part that cannot move for eighteen months. Implementability is a property of the part, not of the saving.
Ranking by difficulty converts a synergy study into an integration plan. Ranking by size produces a headline that the operating team quietly abandons in year one.
The half that price analysis does not cover
Price arbitrage is the visible portion of supply chain value in a deal. The rest sits in the conditions attached to realising it.
Cost to achieve. Requalification, PPAP where applicable, tooling transfer and customer sign-off. In automotive and regulated supply, nine to eighteen months per family before the first dollar lands. Phasing, not the saving itself, is where most synergy cases fail.
Contracts and tooling. Tooling sitting on supplier premises with unclear title. Change of control clauses triggered by the transaction. Minimum volume commitments. Exclusivity that outlives the deal. None of this appears in a spend cube. All of it appears in the first quarter after close.
Concentration risk. Single-source parts with no qualified alternate. Suppliers financially dependent on the target. A sub-tier both entities share without either knowing it. These are valuation inputs, not integration problems.
Working capital. Payment term differentials between the two entities, inventory held against obsolete SKUs, consignment arrangements. Term harmonisation is frequently the fastest cash in the deal and almost never in the model.
The economics of doing the work
A synergy analysis of this kind runs four to six weeks per pairing with a small team. Set that against a synergy line carrying tens of millions of enterprise value.
If the work moves the underwritten number in either direction, it has returned its cost many times over. Moving it down is not a failed diligence. It is a repriced deal, and a repriced deal is a better outcome than an integration plan built on a number that was never real.
The timing argument is equally simple. The combined footprint is cheapest to model before the letter of intent and most expensive to model during integration, when the same analysis has to be done anyway and the price is already fixed.
How Kepler Advanced Analytics approaches this work
Kepler Advanced Analytics runs practices across procurement analytics, supply chain analytics, business operations and M&A advisory, with delivery built around data engineering rather than interview-led benchmarking.
On the transaction side, that includes matched SKU purchasing synergy analysis across two purchasing bases, of the kind described above, delivered in four to six weeks per pairing with the matching logic, normalisation basis and extrapolation assumptions documented for diligence review.
On the portfolio side, the same engine runs at single-company level. Twelve months of AP and PO data becomes a spend cube with price variance by item across suppliers, tail spend exposure, and should-cost gaps built up from material weight, process time and secondary operations rather than taken from the supplier quote. Output is ranked by implementation difficulty. Typical duration is four weeks.
Related capability in the operating layer:
- Multi-ERP spend consolidation, including structured and unstructured procurement data pulled into a single analytical layer. For one of the top players in the North American commercial truck market, a procurement control tower built over roughly a terabyte of multi-source data reduced commodity manager preparation time ahead of supplier negotiations by 80%.
- Network and inventory optimisation. For a US industrial hose and fluid sealing distributor, service level coverage moved from 64% to 99.3% while transportation cost fell 15 to 20% and miles per delivery fell around 40%.
- Category management and OTIF visibility for North American retail distribution.
The common thread is that the number is built from transaction-level data and can be defended at part level, which is the standard a synergy figure has to meet once it enters a valuation.
The question worth asking
Buy-and-build strategies run on the purchasing synergy number. Most of the time it is a percentage nobody has tested against a matched SKU analysis.
When did yours last come from one?