Advertising agency media-account momentum: Largest advertising holding companies are gaining media share, with Publicis leading net new billings

Overview

This data-tracking note examines 2026 year-to-date media account movements using COMvergence data. Goldman Sachs concludes that the largest advertising holding companies are gaining share as demand for scale reinforces a three-player market, with Publicis showing the best net-new-billings outcome and Omnicom remaining preferred alongside it.

Core views

Goldman Sachs introduces a tracker of agency media account wins and losses using COMvergence data for the first seven months of 2026. Its central finding is that the three largest holding companies—Publicis, WPP and Omnicom—are winning business while smaller agency groups have been share donors. The firm views this as evidence that the sector is moving toward a three-player market in which clients value scale. The three leading groups each secured broadly similar gross new billings of about $3 billion year to date: Publicis won $3.201 billion, WPP $2.915 billion and Omnicom $2.908 billion. The differentiation is in losses. Publicis lost $867 million, producing net new billings of $2.334 billion; Omnicom lost $1.707 billion, leaving $1.201 billion of net wins; and WPP lost $2.616 billion, leaving net wins of $299 million. Goldman Sachs therefore identifies Publicis as the clear leader in net new billings, while noting that WPP's positive year-to-date net result is still constructive. The source of new business supports the consolidation argument. Although some large accounts moved among Publicis, WPP and Omnicom, 60% of new billings for the leading groups came from other agency groups. The data recorded 23 account moves above $100 million. Examples include Publicis winning Microsoft's $796 million account from Dentsu, WPP winning JLR's $659 million global business from in-housing, and Omnicom winning Subway's $278 million account from Dentsu. Goldman Sachs also found no cases of large agencies losing major media accounts to in-housing; JLR instead moved in-house business to WPP. Including Havas and Dentsu, account wins totaled about $10 billion through July, only about 40% of the total observed across all of 2025 despite the year being more than halfway complete. The firm cautions that COMvergence primarily captures media wins and losses, rather than all agency work, and can miss undisclosed or smaller moves. It specifically notes that WPP's 2026 creative wins from Henkel and Heineken and Publicis's large CRM-marketing win may not be reflected. Further, billings do not translate uniformly into agency net sales: COMvergence reported $64 billion of WPP billings in 2025 versus approximately $6 billion of WPP Media net sales, implying an average conversion of 9–10%, but Goldman Sachs says this historical average cannot necessarily be applied to current year-to-date net billings. Goldman Sachs says the evidence supports its preference for Buy-rated Publicis and Omnicom. Publicis has a 12-month DCF-based price target of €116, while Omnicom has a 12-month DCF-based price target of $139, raised from $138. For Omnicom, the firm raises 2026 EPS by 3%, primarily for a lower share count, updates interest costs and amortisation in line with guidance, and leaves 2027–2029 EPS unchanged in one description; elsewhere it states changes to 2027–2030 EPS are below 1%. WPP remains Sell-rated with a 12-month target price of 240p, although its positive net new billings are acknowledged. Goldman Sachs identifies the outcome of the Coke review as the important next catalyst.

Analysis framework

Goldman Sachs uses COMvergence's record of media-account wins and losses to compare gross wins, losses and net new billings across agency groups, then examines where the winning accounts came from and highlights moves above $100 million. It interprets the results alongside the limits of media-billings data and separately updates Omnicom earnings assumptions and DCF-based valuation.

Methodology notes

• Industry AnalysisIndustry Concentration Analysis

  Agency-share analysis based on media account wins and losses

  The report compares account movements across major and smaller agency groups to assess whether business is concentrating among the largest holding companies.

• Valuation methodsDCF (Discounted Cash Flow)

  12-month DCF-based price targets for Publicis and Omnicom

  Goldman Sachs values these companies by discounting expected future cash flows to derive its stated 12-month price targets.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Publicis (PUBP.PA)

  Covered agency holding company and Goldman Sachs' preferred name based on the highest year-to-date net new billings.

  Strengths

  Highest net new billings at $2.334bn, driven by the lowest losses among the leading groups.

  Comparison

  Gross wins were similar to WPP and Omnicom at about $3bn, but Publicis had substantially lower losses.

  Risks

  Weaker net new business, weaker advertising demand, generative-AI disruption, consulting competition, margin pressure from reinvestment and value-destructive M&A.

• Omnicom (OMC)

  Covered agency holding company and Goldman Sachs-preferred Buy-rated name.

  Strengths

  $1.201bn of net new billings; 2026 EPS increased 3% mainly for a lower share count.

  Comparison

  Its net new billings trail Publicis but exceed WPP's.

  Risks

  M&A integration execution, weaker marketing spending, generative-AI disruption, FX volatility and weaker advertising-creative performance.

• WPP (WPP.L)

  Covered agency holding company; Goldman Sachs remains Sell-rated despite positive year-to-date net new billings.

  Strengths

  Positive net new billings of $299m and a $659m JLR win from in-housing.

  Weaknesses

  Losses of $2.616bn materially reduced net wins relative to Publicis and Omnicom.

  Comparison

  Gross new billings were similar to the other two largest groups, but WPP's losses were much higher.

  Risks

  A faster-than-expected return to organic growth, further material account wins, asset disposals, additional cost savings, or macro and geopolitical improvement that supports organic growth.

Key data

• Publicis net new billings$2.334bn2026 year to July; $3.201bn of wins less $867m of losses
• Omnicom net new billings$1.201bn2026 year to July; $2.908bn of wins less $1.707bn of losses
• WPP net new billings$299m2026 year to July; $2.915bn of wins less $2.616bn of losses
• Major account moves23Moves with billings above $100m
• Total account winsc.$10bnIncluding Havas and Dentsu through July 2026, about 40% of 2025's overall level
• WPP billings-to-media-net-sales reference$64bn billings versus c.$6bn net sales2025 reference implying average net sales of 9–10% of COMvergence billings
• Omnicom 2026 EPS revision+3%Mainly reflects a lower share count

Impact & implications

The report argues that account-win trends favor the largest holding companies and reinforce a market structure increasingly centered on Publicis, Omnicom and WPP. Publicis's lower losses make its net-new-business position strongest, while Omnicom's model update supports its maintained preference; WPP's positive net wins do not change Goldman Sachs' Sell stance.

Risks

• For Publicis, weaker-than-expected net new business, softer advertising spending, generative-AI-led deflation or disintermediation, consulting competition, heavier reinvestment or value-destructive M&A could challenge the view.
• For WPP, an earlier return to organic growth, further major account wins, asset-disposal proceeds, additional cost savings, or improving macro and geopolitical conditions could challenge Goldman Sachs' Sell view.
• For Omnicom, M&A integration difficulties, reduced or delayed client marketing spending, generative-AI disruption, FX volatility and weaker creative-advertising performance are identified risks.

What to watch

• The outcome of the Coke review, which Goldman Sachs identifies as the next important catalyst.
• Further media-account wins and losses, particularly whether gains continue to come from smaller agency groups.
• Whether broader and non-media agency wins alter the picture provided by COMvergence data.