Agency Holdcos: In-Housing's Limited Threat While Other Factors = Bigger Risks To Growth
Beginning in January 2024, most Madison and Wall research and related data published on this Substack will primarily be available to consulting or advisory clients, or otherwise as part of a paid corporate subscription. Please reach out to brian@madisonandwall.com if you would like to discuss these services or for more information about the new offering.
The World Federation of Advertisers, a global trade group representing “over 150 of the world’s biggest brands” released a report for its members on Wednesday focused on Global Trends in Agency In-Housing, with a headline “In-housing set for rapid and continued growth at major multinationals.”
While I don’t doubt the accuracy of the survey, I think it’s important to put it in some perspective. I’m skeptical of the scale of the impact that in-housing led by the marketers who work with the WFA is having on agency holdcos. At the same time, there are other more pressing long-term concerns that agencies probably should be focused on. Here’s why:
According to the WFA, the report indicates that “66% of brands now have in-house agencies, with 21% actively considering one.” It goes on to state that “The 66% figure represents a 16% rise on the number with these resources in 2020, the last time that this research was carried out.” The survey was based on 45 companies who responded, with what the WFA described as an estimated global ad spend of $60 billion. It’s unclear to what degree the pool of advertisers who responded in the WFA’s survey were representative of the broader industry of marketers.
7% of respondents – presumably meaning 3 – spent more than $50 million on their in-house agencies, while a similar number spent $25-50 million. 33% (or 15 respondents) spent between $5-$25 million while 13% (6) spent $1-5 million and 27% (or 12) spent less than $1 million. Spending levels for 13% of respondents (another 6) were not described in the summary.
What does this mean in context of spending on the broader agency services sector?
Let’s start by estimating how much total spending on in-house activities this group represents.
As even the largest marketers don’t spend more than a few hundred million dollars per year on agency fees, if we assume the top 3 respondents spent $100 million each and further assume the lowest 12 average at the high end of their range, for $1 million each. From there we can calculate that advertisers who responded to the survey only spent around $700 million. Add in something else for the respondents whose spending was not described, and perhaps we are at $800 million on an annual basis. This would equate to just over 1% of this group’s annual media expenditures, which is not a mark of efficiency, but instead is a sign of how little activity is actually in-housed. At these levels, it’s impossible to accomplish much in the way of creative, media and other activities performed by agencies unless one let an AI tool run wild.
For reference, at each of Omnicom and Publicis the top 100 advertisers spend an average of around $75 million per year with those agency groups, which is probably similar for the 500 largest marketers who work with agency holding companies. Each of these marketers probably has a client team averaging around 500 FTE (full time equivalent) staff with many more assigned on a fractional basis to apply specialist expertise.
If the group who responded to the survey represented $60 billion in spending, it would account for just under a quarter of the approximately $250 billion in annual expenditures which went through large agencies globally, as tracked by COMvergence. The respondents here appear to be larger than average clients served by the holdcos. But if we assumed that all small and large advertisers are equally likely to in-source (arguments could be made either way), then multiplying the $800 million figure by 4 would lead us to $3.2 billion in spending. If only two thirds of marketers engaged in any in-housing, perhaps a more appropriate figure would be $2 billion. By contrast, the six biggest agencies generated $55 billion of revenue during 2022.
In other words, even if there has been a significant amount of growth in in-housing and even if there is likely to be more in years to come, it’s unlikely that any spending shifts have accounted for much more than tenths of percentage points of growth in any one year.
So where are the threats to agency holding companies?
To be sure, 2023 was a modest year for agencies. The five broadest globally agencies will probably end up with around 3% growth for the year (the six biggest will end up slightly lower than that given the challenges Dentsu has faced). Although this is somewhat weaker than I would have expected for the group at the beginning of the year (4% felt about right earlier in the year), the biggest drag was likely sector-specific: the significant cuts we saw from tech companies in particular during this “year of efficiency” for much of Silicon Valley.
Looking forward, I can point to several factors that may cause under-performance which are more meaningful than in-housing from large brand owners, including:
Mid-sized agencies
IT services firms
E-commerce and growth of “digital endemic” marketers
Mid-sized agencies have been a real issue over the past decade, as this group of businesses has seemingly outpaced the holdcos over that time. Earlier this year I estimated that in the United States in the years leading up to the pandemic, the overall industry grew by mid-single digit levels while the biggest holdcos were flat with all of the growth accruing to their smaller competitors.
Although this group of agencies has slowed disproportionately through the first three quarters of the year, to the extent they continue to invest more intensively in growth opportunities relative to the holding companies, they could outperform again in the future.
IT services firms and entities focused on marketing-led business transformation are another source of competition, although it’s hard to say how much they impact the holdcos. For example, Accenture reported its fiscal first quarter earnings this week and indicated that its agency-like division Song posted double-digit growth yet again. That unit posted $18 billion in revenue during the company’s fiscal 2023, up by 14% in constant currency terms. However, as Accenture reports revenues that comingle results from different divisions of the company it’s impossible to know what the comparable growth rate is. Alongside Song’s $18 billion, on its earnings release, Accenture highlights that its Cloud business posted $32 billion in revenue, up by 27%, while Industry X saw $7.5 billion in revenue, up by 20% and Security was $7 billion up by 24%. Although the company discloses there is “overlap” between these segments, on a combined basis, if they were additive they posted $64.5 billion in revenue, or more than Accenture’s grand total of $64.1 billion across the whole company, with growth of 22% vs. the actually-reported 8% gain.
Most importantly, I think that anywhere there is e-commerce-related marketing is likely an area that is under-weighted activity for the largest agency groups and a source of under-performance vs. the broader media industry. Companies from all around the world who are primarily dependent on Amazon or Alibaba if in China, and who consequently don’t really need services traditional agencies offer to larger brand owners probably don’t spend much with the holdcos. Chinese manufacturers who buy media on Amazon and who also spend heavily with Meta are similarly not exposed to agency activities for the most part.
Perhaps as important, there are other “digital endemics” – online-born brands whose primary businesses operate primarily on the internet – who, like most businesses, started their businesses with internal marketing operations and who spend much less on agency services per dollar of media expenditures as a result. This latter consideration is generally underappreciated. Most of the marketing services work for these companies is presently in-housed, but the direction of travel for these marketers will likely involve much greater reliance on outsourced agency services as they mature.
Of course, that will only occur so long as agencies invest in themselves in order to evolve the services they offer to better meet the market’s future needs. Omnicom’s recent acquisition of Flywheel was certainly helpful in this regard, but others will need to make more investments (internally or externally) like it as well. They’ll also need to continue to push harder on bundling principal-based media with services, as we’re seeing at Publicis and Omnicom. More focus on data-focused products such as what we’ve seen from Publicis with Epsilon and IPG with Acxiom will help, too – but of course execution will matter as we see with the struggles Dentsu is experiencing despite ownership of Merkle. At the same time, if agencies can continually focus on the ways they can they help marketers drive costs out of their businesses while supporting the pursuit of outcomes (however marketers choose to measure them), they will be well-positioned to see growth that comes closer to total media spending increases in years ahead.