After the deal: four integration styles and how to choose

You have just signed the purchase of another company. On Monday morning the practical question is whether to fold them into your systems, let them run alone, or do something in between. One easy way to answer is by a feeling about culture. This post asks a different question: how much do the two sides need each other, and how much does the other side need to be left alone?

What are the four integration styles?

Haspeslagh and Jemison (1991) propose four ways to integrate after an acquisition. I describe them from the library note on this site, because I have not read the book. The axes are the need for strategic interdependence (how much the two firms must share capabilities and resources for the deal to make sense) and the need for autonomy of the acquired unit (how far it must be left to run itself to keep what was bought). The two glosses in brackets are mine, not the library's (Haspeslagh and Jemison, 1991).

Read from the note: absorption is low autonomy with high interdependence, meaning full merger into the buyer. Preservation is high autonomy with low interdependence, meaning the business keeps operating separately. Symbiosis is both high: two-way learning and gradual transformation. Holding is both low: the relationship is mainly financial (Haspeslagh and Jemison, 1991).

I take two things from this layout. First, no cell is "best": forcing tight integration does not automatically yield more synergy if the other side needs autonomy to keep what you paid for. Second, "holding" is a legitimate choice rather than a failure to integrate; a deal can be mainly a financial investment. Both are inferences from the two-axis layout, not wording from the authors.

Choosing by culture or by the kind of value?

The library note records the pitfall: choosing the style by culture rather than by the kind of value to be transferred (Haspeslagh and Jemison, 1991). This is my reading of it. Culture is the easy answer: "they are too disorganised for us" or "they are so different, leave them alone". Both sound reasonable and both can be wrong if you have not asked what you bought.

The order of questions I would use, built on the two axes:

1. What did we pay to get, and when is it worth anything? A team of engineers, a brand, a distribution network, or cash flow?
2. To collect that value, what must cross the boundary between the two firms: data, processes, people, products? The more that must cross, the higher the need for interdependence.
3. If we interfere with how they work, does what we bought break? If so, the need for autonomy is high.
4. How far apart are the two sides' cultures and processes, and what would it cost to narrow the gap? Culture comes in only here, as a cost of the chosen style, not as the ground for choosing it.

That order is mine. It follows the pitfall the library records and is not quoted from the book.

What do three public deals show?

The library holds three cases with public sources. I use them to test which of the questions above the public documents can answer. This is not a verdict on success: I have no source on what happened after any of the deals.

Microsoft and Nokia (release of 3 September 2013). Microsoft would buy substantially all of Nokia's Devices & Services business for EUR 3.79 billion and pay EUR 1.65 billion more to license Nokia's patents, EUR 5.44 billion in cash in total. It would acquire the design team, the related production facilities and the sales and marketing activities of the business; about 32,000 people were expected to transfer at closing. Microsoft said it aimed to accelerate its share and profit in mobile devices through faster innovation, increased synergies and unified branding and marketing. Nokia keeps its patent portfolio and continues to own the Nokia brand, while Microsoft acquires the Asha and Lumia brands and licenses the Nokia name for existing Nokia phones (Microsoft Corporation, 2013). See also the Microsoft–Nokia case.

On my reading, the announced intent leans towards high interdependence: the release talks about synergies and unified branding, and transfers production and sales people along with design. It does not describe in detail how much autonomy Nokia's side would have after closing, and Microsoft's forward-looking statements list the effectiveness of integrating the business as a risk factor (Microsoft Corporation, 2013). I cannot infer the style actually used from this, still less the result.

Disney and Marvel (release of 31 December 2009). Marvel shareholders receive USD 30 in cash plus about 0.7452 Disney shares per Marvel share, based on Disney's closing price on 31 December. Marvel's assets include a library of over 5,000 characters featured in a variety of media over 70 years, and businesses in licensing, movie production and publishing. Disney said the creative and business potential of the combination is "substantial" (The Walt Disney Company, 2009). See also the Disney case.

Question one, what was bought, is where the release starts. It lists Marvel's assets as a library of characters and businesses in licensing, film and publishing, and Disney's chief executive welcomes Marvel's "talented team". But it does not describe how Disney intended to run Marvel, so it cannot be placed in a cell. I can only frame the question: if the value sits in that team, does heavy-handed interference damage it?

Kraft and Cadbury (Form 8-K, 3 February 2010). Kraft offered 500 pence in cash and 0.1874 Kraft shares per Cadbury ordinary share. If it acquired all the shares, it estimated it would issue no more than about 265 million shares and pay about GBP 7.1 billion in cash. By 2 February 2010 it had received acceptances for about 71.73 per cent of Cadbury's share capital and declared the offer unconditional (Kraft Foods Inc., 2010). See also the Kraft–Cadbury case. I find nothing about integration in the filing. It illustrates the limit of the source itself: price and conditions are on file, the integration style is not.

To analyse a real deal with this framework you need documents from the integration phase, not only the release at announcement.

How would it apply to a hypothetical deal?

A hypothetical example. "Electronics Chain A" has 40 shops in southern Vietnam and deals with "Start-up B", a team of 12 building a city delivery app. The names and numbers are my invention. The table shows how the same deal can lead to four different styles depending on the answers to questions one and two.

The trap in this example is that A finds B "chaotic, with no process" and absorbs it at once. If A bought B for the experimentation speed of a team of 12, forcing it through the approval process of a 40-shop chain may destroy exactly what was bought. Conversely, if A bought B to run the app across the whole chain but "leaves it alone because their culture is different", A will not have the shared data the app needs. Both conclusions are my inference from the two axes; nothing in this post measures them.

This sorting suits a deal with a clear purpose and one dominant answer. It does not suit a deal bought for several reasons at once: then split B into its parts and ask the questions of each, rather than giving the whole deal a single label.

Going deeper: what I read and what I did not

I have not read Haspeslagh and Jemison's book "Managing acquisitions: creating value through corporate renewal" (Free Press, 1991). Everything about the four styles, the two axes and the culture pitfall comes from the library's note on the framework, so I do not know how the authors define the two needs, which examples they use or how they argue the pitfall. The bracketed glosses in the first section and the list of four questions are mine. If you use the framework on a real decision, read the book first.

I read the Microsoft and Disney releases in full and the passages on terms in Kraft's 8-K. The library also lists a Disney source on the 2019 21st Century Fox deal; I did not use it because this post covers only Marvel, Nokia and Cadbury. None of the three documents I read reports results after the deal, so this post does not say that any deal succeeded. A release is written by the buyer at the time of announcement, so the aims in it are expectations, not outcomes.