The Moat Around Every Ad Agency is Client Retention
(This post is a reprint of one of the nine sections that make up the Singular Diligence issue on Omnicom)
Moat is sometimes considered synonymous with “barrier to entry”. Economists like to talk about barriers to entry. Warren Buffett likes to talk about moat. When it comes to investing, “moat” is what matters. Barriers to entry may not matter. Thinking in terms of barriers to entry can frame the question the wrong way.
If you’re thinking about buying shares of Omnicom and holding those shares of stock forever – what matters? Do barriers to entry matter? Does it matter if it’s easy to create one new ad agency or a hundred new ad agencies? No. What matters is the damage any advertising company – whether it’s WPP, Publicis, or a firm that hasn’t been founded yet – can do to Omnicom’s business. How much damage can a new entrant do to Omnicom’s intrinsic value? How much damage can Publicis or WPP do to Omnicom? The answer is almost none. In that sense, the barriers to entry in the advertising industry are low but the moat around each agency is wide. How can that be?
First of all, the historical record is clear that among the global advertising giants we are talking about a stable oligopoly. The best measure of competitive position in the industry is to use relative market share. We simply take media billings – this is not the same as reported revenue – from each of the biggest ad companies and compare them to each other. If one company grows billings faster or slower than the other two – its competitive position has changed in relative terms. Between 2004 and 2014, Omnicom’s position relative to WPP and Publicis didn’t change. Nor did WPP’s relative to Publicis and Omnicom. Nor did Publicis’s position relative to WPP and Omnicom. Not only did they keep the same market share order 1) WPP, 2) Publicis, 3) Omnicom – which is rarer than you’d think over a 10-year span in many industries – they also had remarkably stable size relationships. In 2005, WPP had 45% of the trio’s total billings. In 2010, WPP had 45% of the trio’s combined billings. And in 2014, WPP had 44% of the trio’s combined billings. Likewise, Omnicom had 23% of the trio’s billings in 2005, 22% in 2010, and 23% in 2014. No other industries show as stable relative market shares among the 3 industry leaders as does advertising. Why is this?
Clients almost never leave their ad agency. Customer retention is remarkably close to 100%. New business wins are unimportant to success in any one year at a giant advertising company. The primary relationship for an advertising company is the relationship between a client and its creative agency. The world’s largest advertisers stay with the same advertising holding companies for decades. As part of our research into Omnicom, Quan looked at 97 relationships between marketers and their creative agencies.
I promise you the length of time each marketer has stayed with the same creative agency will surprise you. Let’s look at some of the examples. Wrigley – now a part of Wrigley Mars – used Saatchi & Saatchi as its creative agency from 1954 till 1995. Wrigley left Saatchi because of squabbling within the Saatchi family itself. After leaving Saatchi, Wrigley has been with Omnicom from 1995 through 2015. So, 40 years with Saatchi and then 20 with Omnicom.
Procter & Gamble has historically used Saatchi and Grey. P&G’s relationship with Saatchi dates back to 1921. Its relationship with Grey started in 1956. P&G’s relationship with Grey was costly to Omnicom, because P&G acquired Gillette in 2005. Gillette had been a client of one of Omnicom’s agencies for about 70 years. But, since P&G prefers working with Saatchi and Grey – it moved the Gillette account to Grey in 2013. This is how many big clients are lost. The client merges with a company that is served by a competing agency. Unilever has been an even more loyal client than P&G. Unilever prefers working with J. Walter Thompson and Lowe. It has been with J. Walter Thompson since 1902. So, that relationship is now 114 years old. It’s not the oldest relationship for Unilever. The company started working with Lowe in 1899. So, that relationship is 117 years old. Another example in household products brands is Clorox. Clorox has been with Omnicom’s DDB since 1996. Its prior relationship lasted 75 years. So, Clorox chose an agency in 1921. Then, it switched creative agencies in 1996. So, two moves in close to a hundred years.
Some industries have very high client retention. It seems financial services firms mostly leave agencies due to consolidation. So, one bank buys another and they switch to the bigger bank’s creative agency. Otherwise, the relationships are almost all long ones. State Farm has been with Omnicom’s DDB from 1930 till today. Between 2010 and 2011, State Farm shifted some of its business – especially the car insurance brand – to a different creative agency, but it moved all of that work back to DDB the following year. Allstate has been with Leo Burnett since 1957. Geico has been with the Martin Agency since 1994. All of GEICO’s campaigns you remember were created by The Martin Agency. American Express has been with Ogilvy since 1962. Visa is an interesting example of “moat”. Visa has been an Omnicom client from 1985 through today. However, Visa dropped BBDO in 2005. Omnicom asked Visa to limit its search for a new agency to among Omnicom owned agencies only. Visa agreed. And so the firm selected from pitches made by DDB, TBWA, etc. It didn’t consider moving to an agency owned by WPP or Interpublic or anyone else. In 2005, Visa switched from Omnicom owned BBDO to Omnicom owned TBWA. However, it moved back to BBDO in 2012. So, Omnicom retained Visa as a client despite one of Omnicom’s agencies being fired two different times in that period. Wells Fargo has been an Omnicom client since 1996. Discover Card was an Omnicom client from 1987 to 2006. Then, Discover moved from Omnicom to The Martin Agency. By 2006, The Martin Agency would’ve already been well known for the amazing work it had been doing for GEICO throughout the 1990s and early 2000s.
I don’t want to bore you with almost a hundred different examples of how long one client stays with one creative agency. But, I do think this is the most important fact in this entire issue. So, I encourage you to flip to the “Notes” section of the issue and read the list of relationships and their start dates that Quan prepared for industries ranging from carmakers to transportation companies to electronics brands.
Consolidation is the leading cause of losing a once loyal client. The other reason clients leave is because they are fickle. The same brand keeps switching creative agencies. This is common among troubled brands. A good example is Heineken in the U.S. Heineken is an imported beer that basically has the same positioning as “better beer” brands like Samuel Adams. Heineken originally competed with the big, boring domestic brands like Bud. Over the last 20 years, it’s had to compete with U.S. based craft beers and other imports. In the last 10 years, the brand has been in constant turmoil. From 2005 through 2011, Heineken had 4 different CEOs and 4 different chief marketing officers. Its creative agency from 2003 to 2007 was Publicis, then Wieden & Kennedy for 2008-2009, then Euro RSCG from 2009-2010, then back to Wieden & Kennedy for 2010-2015, and then in 2015 Heineken left Wieden and returned to Publicis. You can see this is not a problem with a particular creative agency. It’s a problem with the Heineken brand.
The other reason clients switch is because they belong to an industry with a lot of cyclicality to brand perception. Restaurants and retailers are the most fickle client group. Still, “fickle” is a relative term. Client retention is still very high compared to other industries. Wal-Mart has used 3 agencies over the last 30 years. Darden (Olive Garden) has been with Grey since 1984. McDonalds has used Leo Burnett and DDB since the 1970s.
There is no customer retention figure generally available for Omnicom or for the industry. But, you can easily estimate the retention rate is above 90% based on the nearly 100 relationships we looked at. You can peruse many of these relationships in the notes. From time to time, articles give client retention figures for a single agency. For example, in 2001 and 2002 and 2003 it was reported that DDB retained 98% or 99% of clients. In 2014, it was reported Grey had a 95% retention rate. An industry wide customer retention rate of 90% would be conservative for key accounts. A figure like 95% may be more realistic.
The most important fact to consider is that agencies don’t try to take business from each other. They try to take business once it is “in review”. In 2006, Omnicom’s CEO said: “We’re invited to new business pitches…we can’t create them…Fortune 100 companies, typically take their time and go through a lot of deliberations before they actually…put an account or an assignment into review.”
In 2007, Omnicom’s CEO (John Wren) again said: “…what goes into review is driven by clients…we can’t cause somebody to put their account in review.”
In other words, the only time one agency takes a Fortune 100 type client from another agency is when the client decides first to put the account in review. First, the client says they may fire the agency and then other agencies pitch for the business. When agencies do pitch for an account in review, they don’t usually compete on price. Traditionally, creative agencies took a 12% commission on billings. Media agencies took a 3% commission on billings. Now, the work tends to be fee based. But, it doesn’t matter. It’s clearly been structured the same way to keep price competition from happening. Omnicom has a very consistent EBIT margin from the 1990s through today. It has a quarter century record of extreme consistency in pricing versus its expenses. It’s basically a cost plus business. We can see with WPP – who reports billings while Omnicom doesn’t – that sales were between 20% and 22% of billings over the last 20 years. Again, it’s a consistent number and it’s not lower than the traditional commission structure the industry used. So, there is no evidence of agencies competing for business by trying to offer lower fees and commissions. The moat around each ad agency is wide because: 1) Customer retention is nearly perfect 2) Pricing is very consistent and 3) The relative market share of the oligopolists in this industry is also very consistent. Without changes in prices or relative market share – a business is basically the same from year-to-year. So, ad agencies mostly report earnings growth and declines based purely on the increases and decreases in ad spending by their existing clients. In fact, what you see in the year-to-year results of an advertising company is basically just that. It’s mostly just a record of what the existing client roster did in terms of their ad spending compared to last year. So, the industry is incredibly stable over full cycles. It is cyclical to the extent that ad spending is cyclical and follows the macro economy.