Net revenue vs gross margin: how D2C changes your P&L

Why revenue can fall while profit rises

Stillfront now puts 46% of its bookings through its own direct-to-consumer platform, up from 39% a year earlier. Its Q2 2026 interim report describes the effect on its accounts in one clause:

The same words appear in its Q1 2026 report. The report states the effect; the mechanism below is our illustration, not Stillfront’s own accounting.

The usual cause is presentation. A studio that reports App Store and Google Play sales gross books the full player price as revenue and puts the 30% store fee in cost of revenue. Electronic Arts does this; its 10-K for the year to 31 March 2026 says mobile platform fees “are reported within cost of revenue”. Playtika’s numbers point the same way. Its CFO said on the Q2 2026 earnings call, according to the Motley Fool transcript, that cost of revenue fell 1.5% year on year, “primarily driven by lower platform fees resulting from the continued growth of our DTC business.” Platform fees can only lower cost of revenue if they sit in it.

On the web, the store fee disappears and a smaller payment cost replaces it. Playtika’s 10-Q puts that cost at “typically 3 to 4%, compared to the 30% platform fee for third party platforms.” If the web sale is also booked gross, revenue stays the same and gross profit rises. If it goes through a merchant of record and you book only what the MoR pays you (the net view), each dollar that moves off the store is reported as less than a dollar of revenue, while more of it stays with you.

EA also shows that one company can present channels differently. It reports console storefront sales net of the storefront’s fee and mobile store sales gross, because it reached different principal-versus-agent conclusions for each contract. How to reach that conclusion for your MoR contract is on merchant of record accounting.

Four ways to book $100 of player spend

The table below is illustrative. It uses round assumptions, not any company’s figures:

• $100 of player spend, excluding VAT and sales tax, which are never revenue.
• A 30% store fee.
• An MoR that keeps 7% of player spend in total.
• Payment service provider (PSP) costs of 4% for processing and related charges, the top of Playtika’s reported range.
• $10 of other cost of revenue (hosting, royalties, player support) in every case.
• No web bonus or discount.

Three readings for the board:

• From (a) to (c) is the Stillfront pattern. Revenue falls 7%, gross profit rises 38% and gross margin rises from 60% to 89%. A studio that books store sales gross and MoR sales net will report exactly this direction.
• From (b) to (c), revenue rises 33%. A studio that books its store sales net, as EA does for consoles, sees the web shop lift revenue and gross profit together.
• Between (c) and (d), the PSP route shows $3 more gross profit but a lower gross margin. Neither is the full picture. The MoR’s 7% pays for tax compliance, fraud screening and chargeback handling. On the PSP route those costs remain, but tax software, fraud tools and the staff who run them often sit in operating expenses, below gross profit. Compare the routes on total cost, which is what the TCO calculator does, and see hidden fees for the lines to include.

If your auditor concludes that you are principal on MoR sales (the view Paddle’s help center takes), column (c) becomes revenue $100, MoR fee $7 in cost of revenue, gross profit $83 and gross margin 83%. Gross profit is identical. Only revenue and the margin percentage change.

What a shift in channel mix does to reported numbers

Take the store gross (a) and web MoR net (c) presentations and move a growing share of the same $100 of player spend to the web shop. Every figure below follows from the assumptions above.

At 40% on the web, revenue is down 2.8% and gross profit up 15% on unchanged player spend. A revenue-only view reads that as a weaker quarter. For scale, Playtika reported 39.3% of revenue through its own channels in Q2 2026, and Stillfront 46% of bookings; the D2C share tracker follows the listed companies each quarter.

Three things in real results can move the numbers further than this table:

• Web incentives. Bonus currency keeps the price and gives the player more items, so revenue per dollar of spend holds. A lower web price cuts player spend for the same items and reduces revenue and gross profit together.
• Platform fees on linked sales. Apple charges 0% on US link-outs today and has asked the court to approve 15%. Since 1 October 2026 Google’s terms have required 20% (10% on the first $1 million) on US purchases made within 24 hours of a link click, a fee whose legality is disputed. Each changes the web column. See Apple’s US rules and Google Play.
• Refunds and chargebacks. On a PSP route they reduce your revenue or add to your costs directly; on an MoR route they depend on the contract. Chargebacks and fraud covers who pays.

Light & Wonder, which runs SciPlay, cites in its Q2 2026 reporting the “higher margin on our direct-to-consumer platform revenue”. Listed publishers are telling investors the same thing in different words; the ones that also book web sales net have to explain the revenue line as well.

KPIs to put in front of the board

Revenue and gross margin percentage both move with presentation. Brief the board on measures that do not.

The KPIs guide covers the operating metrics (conversion, repeat purchase, average order value) that sit under these.

Before the next board pack

1. Get your auditor’s written view on presentation for each channel: each app store, each PC store and the web shop. For an MoR, that means a principal-versus-agent memo on the contract.
2. Restate the last four quarters on a player-spend basis, with a bridge from player spend to revenue to gross profit for each channel.
3. Add gross profit per $100 of player spend and effective take rate to the board pack, next to revenue and gross margin.
4. If you give revenue guidance, show how much of any change comes from channel mix and presentation rather than from player spend.
5. Put the paragraph explaining the revenue effect in the pack before the first quarter in which web share passes 10%, so the board reads it before it sees the revenue line.