Sir Martin Sorrell, the chief executive of WPP, cheerfully compares himself to Napoleon, because he is the same height as the French leader, 5ft 2in. It is perhaps not the ideal comparison because whereas Sir Martin is an empire builder, the French emperor ran into trouble in Russia - but it is not outrageously immodest either.

Two decades after Sir Martin took control of a small manufacturer of wire baskets to create an advertising and now marketing services group, he runs a company with annual revenue of £6.1 billion and profits of £719 million. For all the size of the business, Sir Martin is its beating heart, working with a fearsome energy that leaves anyone near by exhausted. Most chief executives hide behind armies of public relations agencies, many owned by WPP. Not Sir Martin, who calls and e-mails, ready to argue over yesterday's headline while running a multinational at the same time.

WPP has been built through acquisitions - the deals of the 1980s, J Walter Thompson and Ogilvy & Mather being the largest two - nearly killed it when recession followed. Yet, after a restructuring, the restless activity picked up again. Sir Martin sucked in Young & Rubicam in 2000 (which has struggled to generate a return until relatively recently) Tempus, the media buyer, in 2001 (despite trying to wriggle out of it after September 11) and Grey - which has done well - in 2005.

It's great theatre for the financial pages, where few bother to understand what's inside the company; instead they can simply wait for the next big takeover battle to come along, which mercifully it has, in the form of Taylor Nelson Sofres (TNS), the market researcher.

The real problem for Sir Martin, though, is whether the activity down the years is reflected in the share price. Over the past three and five years, WPP shares have underperformed the FTSE 100 and the S&P 500, its nearest US equivalent, although the stock is (just) up. You have to go back ten years to get a beat-the-market result and interestingly, in each of those three periods, WPP does worse than its closest competitor, Omnicom. Nobody outside advertising has heard of Omnicom, which is curious for a communications business but the key point is that the New York company behind the TBWA and DDB agencies doesn't go in for big deals.

Nor is it necessary to be big to be successful in the marketing business. BBH, arguably the most creative advertising agency, is independent; YouGov has done pretty well as a start-up polling firm; Aegis keeps its nose clean as a media buyer. Sir Martin says that his supergroup delivers in the back-office areas, such as property and technology, but a lot of the smaller firms seem able to reach similar profit margins.

WPP's margin is a shade over 15 per cent, a bit higher that most rivals, except Publicis. However, it is not dramatically higher - two percentage points back is Aegis, and TNS is four points lower. The latter's position, though, reflects the more labour-intensive nature of market research, in which many projects are conducted for just the one client. WPP's own market research arm operates on a par with TNS, demonstrating that what matters is not size per se, but size in a subsector of the marketing business. Quizzing housewives in Redcar is not the same as shooting an ad in Argentina.

It is easy to conclude that Sir Martin wants TNS now because the latter has put itself in play by unveiling a merger with Germany's GfK. Certainly, the hypercompetitive WPP chief likes to disrupt rivals as much as to build his own business, but there are other reasons why it is interesting.

Market research has historically been more resilient in an economic downturn, which is expected to start in 2009 once the short-term lifts of the US presidential election, the Olympics and Euro 2008 are over. Sir Martin voices some doubts about 2009, and so more market research may help to avoid a nightmare on Farm Street, where WPP has its London headquarters.

Nevertheless, Sir Martin tells all who will listen that he is often prepared to walk away, although it is hard to think of an example in the public domain. Nobody expects him to abandon TNS easily, as implied by all the noise from the WPP camp. We hear about how the merger with GfK won't work, because TNS management struggled with its last big American buy, NFO, or because their 15 per cent margin target is unachievable. Unfortunately, although the talk is everywhere, it is hardly dampening TNS's share price, off 2p to 257p yeterday. If Sir Martin wants to do another deal, and it has been three years since the last one, he will have to pay up.

Yet, for all the nagging doubts, it is impossible not to admire Sir Martin. In an era in which most British media companies are struggling because they failed to globalise, WPP stands out and it has been led by the same man for more than two decades. And although larger size does not lead to significantly better performance (despite one percentage point of margin being £60 million of profit), WPP's size is not a disadvantage either. Its actual agencies operate in small, autonomous units. The share price could do better, but the battle for TNS is no Waterloo.

Times Recommends

• Gazprom brands Ukraine 'a barbarian' in gas row
• Services sector raises fears over UK jobless
• Retailers surge as profits survive sales plunge