The Arena Group
in partnership with Mile AI Prebid Server
Performance Report
May 4 – May 18, 2026
Prepared by Mile
Download PDF
Executive summary

Mile is delivering measurable incremental yield across The Arena Group's top revenue properties — and the trajectory is accelerating.

Across fifteen days of production data on The Arena Group's top 7 revenue properties, Mile bidder contributed 10.7% of total bidder revenue across the reporting window, and following the May 15 algorithm update ranks #2 across The Arena Group's network in aggregate — behind only Kargo, and ahead of Teads, IX, and every other demand partner. The data tells a story of compounding momentum — Mile's revenue share began climbing from the May 10–11 baseline, then accelerated after the May 15 algorithm update to 12.1%, sustained through the close of the reporting window. Ten demand partners are now live on Mile's server-side container, delivering same-partner CPM premiums of +15% to +24% versus their client-side adapters. Two more partners — IX and Criteo — have server-side endpoints pending activation; once live, Mile's projected contribution rises to ~20% of total Arena programmatic revenue.

Mile revenue contribution
12.1%
Top 7 properties, May 15–17
Up from 10.2% pre-update
Same-partner CPM premium
+19.0%
Revenue-weighted across all live
Mile SS partners with CS counterparts
Mile rank on top properties
#2
Network-wide post-May 15
Behind Kargo, ahead of Teads & IX
Live SS demand partners
10
Active on Mile's container
Two more pending: IX, Criteo

Headline wins

What's working on Arena

Mile's contribution is consistent, measurable, and improving over time. Four wins from the last fifteen days illustrate where the partnership is generating value today.

01

Mile is the top bidder on autoblog.com across every metric

On autoblog.com, Mile holds the #1 position by Revenue ($9.3k), Average CPM ($5.15), and Request RPM (0.32). Mile contributes 19% of total bidder revenue on this property — nearly double Mile's share on the rest of the Arena portfolio. autoblog represents the partnership's mature state: the optimization strategies are running at full effect.

02

Server-side path consistently lifts partner pricing

For every partner currently live on Mile's server-side path with a comparable client-side adapter — pubmatic, rubicon, sovrn, triplelift — the same partner clears at a higher CPM through Mile than through their own client-side seat on the identical inventory. Triplelift +24%, pubmatic +21%, rubicon +16%. Direct evidence Mile's algorithms generate genuine incremental value.

03

Mile's algorithms are continuously learning

Mile's revenue share began rising from a steady ~10% baseline starting May 10–11, then accelerated after the May 15 algorithm update to 12.1% — sustained through the close of the reporting window. CPM rose on every top property (+1% to +25%) and Request RPM rose 24–56% on six. This is a compounding optimization cycle, not a one-time lift.

04

Mile is the #2 demand partner across The Arena Group's network

Following the May 15 algorithm update, Mile ranks #2 in aggregate revenue across The Arena Group's top 7 properties — behind Kargo and ahead of Teads, IX, triplelift, and every other demand partner. Mile holds #1 outright on autoblog.com and #2 on mensjournal.com and trending.thespun.com. The network-level #2 position is the direct result of the compounding optimization gains across the reporting window.

Mile's revenue contribution by Arena property
Mile bidder's share of total programmatic revenue, by property. autoblog.com represents the mature state of the partnership — what the rest of the portfolio is moving toward.

Domain performance

Mile performance across The Arena Group's top 7 properties

Seven properties account for 90% of The Arena Group's programmatic revenue. Mile is active on all seven — and outprices the peer average on every single one. The range of outcomes across domains tells a clear story: autoblog.com is the mature state, mensjournal.com is the fastest path to replicating it, and the May 15 update delivered meaningful gains across the tail.

Mile CPM vs peer average by domain
Mile's average CPM vs the peer bidder average on the same inventory. Mile clears above the peer average on every property — the premium reflects Mile's optimization layer improving yield on each impression Mile wins.

Reading the domain spread

autoblog.com is the ceiling. At +37% above peer average CPM, 19% revenue share, and #1 rank across every metric, autoblog represents what the partnership looks like when Mile's algorithms have had time to optimize on a property. Win rate here is 6.1% — the highest of any domain — meaning Mile is not just pricing higher but winning more often too.

mensjournal.com is the next autoblog. Mile is already #2 by revenue on mensjournal — the only other domain outside autoblog where Mile has broken into the top two. Bid rate is the highest of any property at 36.2%, and CPM is competitive. The gap to close is share: 9.9% now versus autoblog's 19%. The optimization levers are working; this domain needs more volume.

Pre/post May 15 share and CPM change by domain
Mile's revenue share and CPM before and after the May 15 algorithm update, by property. Share gains and CPM lifts are broad-based — six of seven properties improved on at least one dimension.

Where the May 15 update hit hardest

paradepets.com: +25% CPM lift (from $2.73 to $3.41) with share up +3.3pp. Paradepets was Mile's weakest property pre-update; the May 15 retrain appears to have found pricing signals that weren't being captured. Worth watching whether this holds over the next reporting window.

trending.thespun.com: +4.7pp share gain — the largest share increase of any property. CPM moved modestly (+1%), meaning the gain was driven by increased bid density rather than pricing. Mile is now at 15.5% share on thespun, up from 10.8%.

athlonsports.com: +14% CPM lift from a low base ($2.88 → $3.29). Share up +1.3pp. athlonsports has the lowest bid rate of any domain (22.2%) — there is likely additional request volume Mile can participate in that it currently doesn't.

autoblog.com: share dipped -1.1pp despite a +4% CPM improvement. This is consistent with a publisher-side traffic pattern on autoblog rather than an optimization regression — Mile's per-impression metrics improved while total domain volume contracted.

Domain scorecard — two periods. May 4–14 is the pre-update baseline. May 15–17 is post-algorithm update. Rank is by Mile revenue vs all other client-side bidders on that property. CPM vs peers compares Mile's average CPM to the peer bidder average on the same inventory.
Property May 4–14 May 15–17 Share Δ
Share CPM Rank Share CPM Rank
autoblog.com 19.3% $5.02 #1 18.3% $5.21 #1 −1.1pp
mensjournal.com 9.5% $3.67 #2 12.5% $3.83 #2 +3.0pp
thestreet.com 9.8% $3.33 #5 11.6% $3.74 #3 +1.8pp
trending.thespun.com 10.8% $2.28 #4 15.5% $2.30 #2 +4.7pp
parade.com 8.7% $2.88 #4 10.4% $3.09 #3 +1.7pp
athlonsports.com 8.7% $2.88 #4 10.1% $3.29 #3 +1.3pp
paradepets.com 6.8% $2.73 #4 10.1% $3.41 #3 +3.3pp

Incrementality validated

How we know Mile is adding yield that wasn't there before

Mile's commitment to Arena was incrementality — measurable lift in programmatic yield from the same demand partners, delivered through a managed server-side optimization layer. Three independent signals in the data validate that this is happening.

1. Same partner, higher prices through Mile

For every demand partner currently live on Mile's server-side path on Arena (with a comparable client-side adapter for direct comparison), the partner's average CPM through Mile is higher than the same partner's CPM through their direct client-side adapter — on the same property, the same time window, the same inventory. The premium isn't marginal: it ranges from +1.3% (sovrn, where Mile is still building scale) to +23.6% (triplelift) with a revenue-weighted average of +19%.

This is the cleanest evidence Mile's optimization adds value rather than substituting for client-side fill. The same buyer, bidding on the same impression, pays more when Mile mediates the auction.

Same-bidder CPM premium server-side vs client-side
Average CPM for the same demand partner through Mile's server-side path vs. that partner's own client-side seat. Identical inventory, different optimization layer.

2. A trajectory of compounding gains

The day-over-day trend tells the most important story. May 4–9 establishes a steady ~10% baseline. Starting May 10–11, Mile's share begins climbing — May 14 hits 11.5%. The May 15 algorithm update then accelerates the curve: May 15 reaches 12.7%, May 16 holds 12.0%, May 17 sustains at 11.7%. The shape of the curve matters more than any single day's number — Mile's contribution is on an upward trajectory, with optimization cycles compounding rather than producing one-time spikes.

Mile daily revenue share with momentum framing
Mile's day-over-day share of total bidder revenue across Arena's top 7 properties. Baseline phase, build-up phase, and accelerated phase shown by point color. May 18 is partial-day data.

3. Compounding optimization across the portfolio

The May 15 algorithm update — a model retrain and floor optimization on Mile's side — lifted Mile's revenue share broadly across the portfolio in three days. Six of seven top properties improved, with no single property disproportionately driving the result. This is what continuous optimization looks like in production: the algorithms learn, the yield compounds, and Arena sees the result in the daily data.

Mile share across three phases by property
Mile's revenue share across three phases of the reporting window. The May 10–14 build-up reflects ongoing optimization gains; May 15–17 shows the acceleration after the most recent algorithm deploy.
Mile CPM and bid density growth
Underlying metric movement: Mile's CPM and bid density (Request RPM) on each top property, comparing the baseline period to the accelerated phase.

Partner unlock — action required

IX and Criteo's server-side endpoints are not yet live

Two of The Arena Group's most significant demand partners — Index Exchange and Criteo — have not yet activated Mile's Prebid Server endpoint on their side. They appear in Mile's configuration but are showing zero server-side revenue and zero bid activity across the entire reporting window. They are not transacting through Mile's server-side container today.

Once activated, IX and Criteo together are projected to add approximately +8 percentage points to Mile's revenue share on Arena's top properties, bringing Mile's total contribution to ~20% of programmatic yield.

+8 pp
Projected lift to Mile revenue share once IX and Criteo activate, based on observed CPM premium pattern from existing live partners.
~$3.5k
IX's current client-side daily revenue on Arena top 7 — the volume waiting to route through Mile's optimization layer.
~20%
Mile's projected total revenue contribution post-activation, up from 12.1% today.
IX and Criteo opportunity sizing
Mile's current revenue share vs. projected share post IX and Criteo activation. The gap represents incremental yield waiting on partner go-live.
What's needed from The Arena Group

No seat-level routing, no configuration changes on your side. The activation requires that IX and Criteo simply turn on Mile's Prebid Server endpoint on their end — Mile's infrastructure is already provisioned and ready. The partner-side action is a one-step enablement.

The most direct path: have your Arena partner-management team request IX and Criteo to activate Mile's Prebid Server endpoint for The Arena Group inventory. Once that's done on the partner side, Mile begins routing immediately and the lift will be visible in the next reporting window.


Partner roster

Demand partners on Mile's server-side container

Mile's server-side container currently routes ten demand partners with measurable revenue contribution on The Arena Group's top properties over the 15-day reporting window. Two more are configured but not yet transacting.

Reading this table. Each row is one SSP/demand partner. SS revenue (15 days) is the partner's total revenue routed through Mile's server-side container on Arena's top 7 properties over May 4–18. Avg SS CPM is the partner's average eCPM through Mile (when they bid). Client-side counterpart indicates whether this partner also runs as a direct client-side adapter on Arena — when "yes," we can measure the CPM premium Mile delivers; when "no," every dollar this partner generates through Mile is incremental coverage with no cannibalization risk.
Partner (SSP) Status SS revenue (15 days) Avg SS CPM Client-side counterpart CPM premium (SS vs CS)
Live partners — measurable CPM premium vs client-side
pubmatic Live $14,412 $3.84 Yes +20.5%
rubicon Live $7,121 $3.35 Yes +15.7%
triplelift Live $4,241 $3.50 Yes +23.6%
sovrn Live $1,070 $2.59 Yes +1.3%
Live partners — server-side only (incremental coverage, no client-side risk)
openx Live $13,484 $3.25 No Bidder not active on client side
medianet Live $5,756 $2.46 No Bidder not active on client side
appnexus Live $3,269 $3.40 No Bidder not active on client side
unruly Live $2,522 $2.87 No Bidder not active on client side
smartadserver Live $317 $2.04 No Bidder not active on client side
sharethrough Live $73 $1.95 No Bidder not active on client side
Configured but not yet transacting — server-side endpoint not active on partner side
Index Exchange (IX) Pending $0 Yes (heavy CS volume) Awaiting partner activation
Criteo Pending $0 Yes Awaiting partner activation

Roadmap

What's next on Mile's side

Mile's optimization investment on The Arena Group continues in two parallel tracks: ramping the partner roster to its full configured state, and continuing the algorithm improvements that drove the May 15 acceleration.

Partner roster completion

IX and Criteo activation is the most consequential near-term action — both are configured on Mile's side and waiting on partner activation of Mile's Prebid Server endpoint. Beyond those two, Mile is actively evaluating Arena traffic to identify additional demand sources to add server-side, both as new partners and as share expansion for partners currently routing only a fraction of their volume through Mile.

Continued algorithm investment

The May 15 deploy is one step in an ongoing optimization cycle. Mile's team has additional improvements in the pipeline — model retraining on the expanded data corpus, floor model refinement on the lower-revenue properties where Mile's share is still below portfolio average, and bid enrichment work that compounds on top of May 15 gains. The next deploy is targeted for early June.

Measurement and reporting cadence

Mile will continue producing this performance report on a biweekly cadence. Any algorithm change deployed in the next window will be measured against the May 15–17 baseline so the incremental contribution remains unambiguous. If The Arena Group would benefit from an alternative cadence or a specific metric in the report, let us know.