OpenPlatforms & Dealsthrough 2029
35–43%
Most likely outcome
Strategic Sale / Absorption — Bridgepoint exits MiQ to a strategic, ending independence
MiQ, the independent programmatic trading desk, and its CTV/ATV bet
Read the full research →01The Outcomes — 4 scenarios, priced by the research
Strategic Sale / Absorption — Bridgepoint exits MiQ to a strategic, ending independence35–43%
The modal terminal state. Bridgepoint's Sept-2022 debt-funded buyout puts MiQ in a 2026-2028 forced-transaction window; with the post-TTD-derating IPO door shut, the rational route is a STRATEGIC trade sale: to a holdco via the Dentsu license-then-buy on-ramp, a challenger network (Stagwell archetype), or an SSP/curation/data player wanting the owned ~90M-HH CTV/ATV footprint + 15-year Sigma corpus. Independence ends; brand may persist as a unit; trader labor partly rationalized; the Sigma re-rate narrative reads partly as exit dressing. This bucket is the STRATEGIC-SALE slice ONLY — sponsor-to-sponsor secondaries and continuation vehicles, which preserve the standalone trajectory, are booked in 'Persisting Independent.'
How this number was derived
Decomposition, reconciled post-cross-exam. P(Bridgepoint pursues exit by 2029 | 2022 debt-funded buyout, ECI 6.1x) ≈0.82 × P(route is a STRATEGIC sale in-window, NOT IPO and NOT a secondary/continuation preserving standalone | thin IPO market, secondaries 25-30% of exits) ≈0.50 ≈0.41, trimmed to 35-43% for slippage past 2029. Cross-exam DEFEATED the contrarian's 42-52% (its 0.65 sale-vs-IPO factor double-counted secondaries/continuations as 'sales'; four members re-routed that mass out). Base-rate/driver/actor members converged ~30-45% on the sale-specific slice; geo-mean-of-odds of reconciled slice estimates (0.41,0.34,0.45,0.44) ≈0.41. Modal but below 50% precisely because non-sale exits sit in the next bucket.
Drivers
- Bridgepoint PE clock: Sept-2022 debt-funded buyout → 2026-2028 window (doc 07); ECI minority already exited 6.1x, confirming the asset trades
- Shut ad-tech IPO window post-2025 TTD ~68% derating pushes exit toward trade sale (doc 06)
- Owned, separable 90M-HH CTV/ATV footprint + Sigma corpus = the scarce acquirable asset (docs 04, 09, 12)
- Dentsu license-then-buy on-ramp as a natural acquirer path (doc 19)
- M&A roll-up (Pathlabs, Adsmovil) reads as build-to-sell packaging (doc 08)
Signals to watch
- Banker mandate, dataroom, or 'exploring strategic options' report on MiQ (Digiday/Axios/FT), Q4 2026 onward — highest-information signal
- Dentsu deepening from license to equity/option (doc 19)
- Founder/CEO equity-rollover or retention language
- Messaging shifts from 'largest independent' to platform-asset value
- A comparable PE-held intermediary trades to a strategic in 2026-2027, resetting the comp
Horizon · Through 2029 (exit window 2026-2028)
Persisting Independent — Recap, Secondary Buyout, or Continuation; Aggregator Survives, Margins Bleed24–32%
Co-modal with the sale. MiQ remains an independent going concern through 2029 via a liquidity event that is NOT a strategic sale — secondary buyout, continuation fund, or recap — OR Bridgepoint holds past 2029. It keeps operating as 'the largest independent programmatic partner': the neutral harmonizer holdcos fear and need, the can't-in-house partner for mid-market/indie agencies (Pathlabs MEP), owner of the 90M-HH CTV/ATV footprint. Margins stay services-shaped (~9-15%) and erode slowly under SPO/curation/TTD pressure; Sigma is a real efficiency layer but NOT a disclosed software-margin business. Survival-as-independent and margin-bleed are the SAME world here; counterparties cooperate because TTD keeps MiQ as re-routable non-Certified spend and suppliers keep selling to a scaled non-exclusive buyer.
How this number was derived
Base rate + exit-route decomposition. Of ~8-10 squeezed independents, ~3-4 stayed durable past their inflection → ~35% persistence base rate; conditioned by P(Bridgepoint LETS it stay independent | survives) ≈0.6 given a 7yr-held debt-funded buyout, PLUS the secondary/continuation share (~0.80 exit × ~0.30 non-sale-route ≈0.24 of mass) that contrarian/driver members mis-routed into 'Acquired.' Lands 24-32%. Cross-exam corrected the driver member's 38-46% DOWN (over-weighted smooth continuity; never multiplied by the conditional that the clock lets it stay independent). Geo-mean-of-odds of reconciled persistence estimates (0.26,0.40-shaved,0.22,0.24,0.26) ≈0.27. Co-modal with the sale — the disagreement over which leads is itself signal on contingency.
Drivers
- Defeats the transparency strand (disclosed-by-birth) and in-housing strand (plateaued ~25% full / ~51% hybrid; skills bind ~3/4 of advertisers; mid-market can't in-house — doc 17)
- Multi-DSP optionality: NOT a TTD Certified Service Partner, free to re-route (doc 20)
- Suppliers' dominant strategy is to keep selling to a scaled non-exclusive buyer (docs 13, 14)
- Secondary buyouts/continuation funds a rising share of PE exits in a thin-IPO market
- CTV/ATV linear-migration tailwind keeps demand growing (doc 10)
Signals to watch
- Bridgepoint announces a secondary sale to another sponsor or a continuation vehicle (this scenario's marker)
- Headcount holds ~1,900 or grows through 2027 (Revelio); no sale process
- No disclosed SaaS revenue line emerges (persistence WITHOUT software transformation)
- Pathlabs MEP / Adsmovil keep posting growth; mid-market retention holds
- Hybrid in-housing share stays dominant vs. full-in-housing breaking out in IAB/IPA data
Horizon · Through 2029 (persistence read extending beyond)
Erosion / Distress — Agentic Commoditization + Supplier Seam Break → Margin Collapse or Distressed Sale15–22%
The bear tail the corpus names as the ONE threat MiQ does not defeat (docs 12, 17). Agentic AI commoditizes managed-service trader labor FASTER than Sigma converts MiQ into a defensible software/data layer, AND/OR a discrete supplier seam breaks (Samba/Titan/Experian acquired by a rival, or Samsung/LG/Roku pull ACR in-house — docs 04, 09, 14), repricing the rented moat. Single-platform agents (TTD's, Amazon's, Yahoo's) plus SPO/curation collapse the contestable middle; performance claims have no independent validation; Pemberton leverage on a ~9-11% margin amplifies any compression. Sharpened by the PE clock, which converts a shopped-but-unsold process into a distressed mark — the MediaMath path, though more likely a haircut sale than a zero given MiQ's profitability/scale.
How this number was derived
Reference-class + correlated-conjunction. Of ~7-10 desk/intermediary lineages, ~1-2 ended in distress/failure (MediaMath $1B→0, Sizmek) → ~12-20% base rate. Council conceded (vs. actor-strategy) the chained factors — agentic erosion, Sigma non-conversion, tip-to-distress — are POSITIVELY correlated (one weak-software root cause), so 0.55³≈0.17 UNDERSTATES the joint tail; floor nudged up. Cross-exam DEFEATED the discontinuity member's standalone 20-30% supplier-shock (unanchored ~7%/yr hazard; Xandr was a re-routable rail not a fatal data defection; doc 09 'aggregator re-routes') — folded in here as a contributor, not its own outcome. Capped below ~22% as MiQ is profitable/scaled (unlike MediaMath) — distress more likely a haircut sale than a zero. Geo-mean-of-odds (0.17,0.20,0.12,0.19,0.29) ≈0.18.
Drivers
- Agentic commoditization of trader labor — the undefeated bear strand (doc 17); Sigma is itself the threat to the labor funding most revenue (doc 10)
- Rented data + acute supplier-concentration; Samba/Experian/Titan also sell the same signal to all and to rivals (docs 09, 14)
- No independent (Forrester/Gartner/G2) validation of performance/agentic claims (doc 16)
- Pemberton debt on a thin-margin base converts compression into distress (doc 01)
- A shopped-but-unsold process converting forced-exit pressure into a distressed mark; failure factors positively correlated
Signals to watch
- A named supplier (Samba/Titan/Experian) acquired by TTD/Amazon/a walled garden, OR Samsung/LG/Roku restricting third-party ACR at scale
- Sigma Trading Agent staying 'announced-not-shipped' past 2027 (doc 12)
- Trader-org layoffs framed as 'AI efficiency' NOT offset by software revenue; visible margin compression
- A covenant breach, refinancing, or down-round/distressed-secondary mark on the Pemberton debt (doc 01)
- A leaked failed sale process ('shopped, no buyer')
Horizon · Through 2029 (acute 2027-2029 for the agentic leg)
Software Re-Rating — Sigma Becomes a Real Margin Story (bull)10–16%
The equity story works: Sigma moves MiQ materially from labor toward software/data margins BEFORE agentic AI makes the human layer redundant. A disclosed or de-facto platform-revenue mix emerges; the Trading Agent scales agentic buying across the book on the 15-year proprietary corpus + neutral cross-platform position (the one moat single-platform agents lack); margins climb off the ~9-11% floor toward high-teens/twenties; MiQ re-rates as a platform. The least common terminal state for a services intermediary — and even where it works, it most plausibly raises the EXIT multiple rather than guaranteeing enduring independence.
How this number was derived
Decomposition — the council's ONE point of full agreement; all five members independently landed 8-18% by separate decompositions, itself strong corroboration. P(Sigma converts to a disclosed/validated software-margin line by 2029 | announced-not-shipped, NO Forrester/Gartner/G2, doc 16) ≈0.30-0.35 × P(it durably re-rates MiQ rather than merely lifting the exit multiple | even success often cashes out as a higher acquisition price) ≈0.45 ≈0.14, set 10-16%. Cross-exam gave NO reason to move it; only the top was trimmed after the panel reinforced every Sigma metric is single-source vendor-voice and PPC Land flags $2.22/$1 covers the OLD platform (docs 02, 12, 16). Geo-mean-of-odds (0.15,0.10,0.16,0.13,0.135) ≈0.13.
Drivers
- Sigma is real and live (June 2025): ~600 feeds, 2.5PB/day, agentic Trading Agent on 15yr proprietary data (docs 02, 12)
- Proprietary corpus + neutral cross-platform position is a moat single-platform agents lack (doc 12)
- PE incentive aligned: the equity story REQUIRES the labor-to-software shift before exit (docs 07, 10)
- Commerce/retail-media (~15.6% of ad revenue) + TV-data footprint feed Sigma differentiated inputs (docs 04, 11)
Signals to watch
- A DISCLOSED Sigma software/platform revenue line or ARR (none exists as of June 2026, doc 07) — the single clearest tell
- First independent validation: Forrester Wave, Gartner MQ, G2, or candid named buy-side testimony (absent today, doc 16)
- Group margin climbing out of the ~10% band toward high-teens
- Trading Agent moving from 'announced/~6,000 campaigns' to verified at-scale autonomous buying, externally benchmarked (doc 12)
- A marquee deal where a holdco runs ON Sigma as software, not as MiQ-managed service (Dentsu, doc 19)
Horizon · Through 2029
Probabilities are the research's own scenario bands, priced as outcomes. The full argument — history, current state, drivers, and sources — lives in the corpus: read the Future Trajectory chapter.
