# Food-Delivery Ads: The Auction, $1B Revenue, Position Bias

Lucas Moreau · August 31, 2026

> Food-Delivery Ads: The Auction, $1B Revenue, Position Bias. The true cost of a sponsored order extends far beyond the advertised take...

| Takeaway | Detail |
| --- | --- |
| Platform ad credits lower initial campaign risk | New sellers launching Sponsored Products within 30 days of listing their first buyable ASIN qualify for up to $1,000 USD in ad credits |
| Sponsored inventory drives a massive share of total marketplace sales | Amazon's recommendation engine and sponsored placements are estimated to influence 35% of the company's total sales |
| Contextual messaging outperforms standard display formats | Contextual in-app messages deliver a 5.6× higher purchase conversion rate compared to standard display placements |
| Marketplace scale amplifies advertising revenue potential | Amazon's 2024 net sales reached $574.8 billion, with over $200 billion annually influenced by recommendation and sponsored placement systems |

The true cost of a sponsored order extends far beyond the advertised take rate. When you factor in the cost-per-click divided by a decayed click-through rate, most campaigns flagged as profitable in vendor dashboards are actually cannibalizing free organic orders. The auction mechanics force merchants into exact-match bidding without negative keyword support, stripping away traditional optimization levers.

As platforms deploy shoppable video, AI-powered conversational search, and viewport-verified impression tracking, the financial architecture grows increasingly opaque. Merchants must recognize that position bias and inflated acquisition costs are systematically eroding margins, turning what appears to be growth into a subsidized loss of previously guaranteed traffic.

The reason the lift rarely clears 1.4x lies in the CTR decay curve, which operates as a position-bias artifact rather than a relevance signal. In food-app recommender systems, top organic cards capture roughly 5–7% click-through rates. When sponsored slots are injected at positions two through four, they decay to 2–3% CTR—a 40–60% drop driven almost entirely by scroll behavior and thumb-zone fatigue, not by menu quality or rating velocity. This decay means the auction captures demand that would have organically converted lower in the feed, re-ranking existing intent rather than generating new discovery. The true incremental lift typically sits between 1.1x and 1.3x, which falls short of the ~1.4x break-even once the 15–20% take rate is charged on top of the CPC. The 15–20% band remains the anchor because it represents the baseline commission across DoorDash's basic tier, Uber Eats' Plus plan, and Grubhub's standard marketplace fee, and it compounds against every ad-driven conversion without discounting.

![Food-Delivery Ads](https://static.mm-ais.com/article-images-ai/food-delivery-ads-the-auction-1b-revenue-ai-f526e643.jpg)

## The Auction Under the Menu

The actionable takeaway is mechanical: treat sponsored slots as a liquidity tool, not a growth engine. Run a geo-holdout or day-part holdout to measure actual incremental lift, then compare it against the 1.4x threshold derived from your specific contribution margin. If the holdout shows 1.1x or 1.2x lift, redirect the budget toward menu conversion optimization, prep-time reduction, and rating velocity—levers that shift the organic relevance score upward without triggering the double-charge. The auction does not create demand; it borrows it at a premium, and the math only works when the borrowed volume exceeds the 1.4x break-even floor.

The mechanics of this decay are predictable in recommender systems. Joachims et al.'s click-model research, examining counterfactual learning-to-rank work, demonstrates that top-position clicks are 3-5x bottom-position clicks independent of relevance. Sponsored placements at mid-page slots suffer CTR collapse even for well-matched restaurants because position bias dominates user attention; the algorithmic advantage of a sponsored slot is often negated by the natural drop-off in engagement as scroll depth increases. Consequently, the effective conversion value per impression drops sharply unless the placement guarantees top-of-feed visibility, which drives CPC costs into unprofitable territory for most merchants.

Uber Eats' published ads materials from Uber Advertising's 2023-2024 merchant case studies report sponsored-listing ROAS claims of 3-6x, but these figures count attributed orders, not incremental ones. Per the attribution methodology Uber itself describes as last-touch, any order placed within the attribution window receives credit to the ad regardless of whether the user would have converted organically. This inflates perceived efficiency while masking the true marginal gain. Grubhub's promoted-listings documentation similarly states promoted restaurants appear in 'top placement' slots with CPC pricing, yet notes Grubhub's standard marketplace commission of roughly 20% stacks beneath the ad cost, compounding the expense on every captured transaction without guaranteeing net margin improvement.

| Metric | Organic Baseline | Sponsored Injection | Impact on Lift |
| --- | --- | --- | --- |
| Position Bias Decay | 5–7% CTR (pos 1) | 2–3% CTR (pos 2–4) | 40–60% drop from scroll behavior |
| Commission Layer | 15–20% take rate | 15–20% take rate + CPC | Double-charge structure |
| $28 AOV Cost | $4.76 commission | $4.76 + $0.30–$1.50 CPC | Pure incremental friction |
| Break-Even Threshold | N/A | ~1.4x organic lift | Required to offset combined costs |
| Holdout Rule | Spend on levers | Skip auction if | Preserve contribution margin |

Academic marketplace-ads research provides the strongest evidence that auction-based placement captures demand organic ranking would have delivered anyway. Ghosh and Mahdian's work on sponsored-search auctions highlights how bidding wars distort relevance signals, while Blake, Nosko, and Tadelis's eBay field experiment shows ~98% of branded ad clicks were incremental-free, meaning the vast majority of traffic would have occurred without paid intervention. In food delivery, where search intent is often low and discovery is driven by proximity and rating velocity, sponsored slots frequently cannibalize organic impressions rather than expanding total category volume. The incremental lift typically lands between 1.1x and 1.3x, failing to clear the 1.4x hurdle required to offset the combined CPC and commission drag.

![The Auction Under the Menu — Food-Delivery Ads](https://static.mm-ais.com/article-images-ai/food-delivery-ads-the-auction-1b-revenue-ai-fbe7af4d.jpg)

## The Numbers on Record

This dynamic is no longer isolated to US incumbents. Deliveroo's and Just Eat Takeaway's investor communications note media and advertising revenue as a growing segment in 2024-2025, confirming the sponsored-slot model is now standard across European and US food marketplaces. As platforms diversify revenue streams away from pure commission dependency, the pressure to monetize discovery infrastructure intensifies. Merchants must treat sponsored inventory as a tax on existing demand rather than a growth lever; only holdout-measured incremental lifts can validate spend, and absent that proof, capital should flow to organic ranking levers like menu conversion optimization and rating velocity management.

When you isolate the four primary growth vectors available to a 2026 marketplace merchant, the economics of sponsored inventory immediately diverge from organic discovery. Sponsored slots operate on a dual-charge model: a per-click auction fee layered atop the platform’s standard 15–20% take rate. Organic ranking optimization carries zero direct cost but requires sustained velocity in rating signals and prep-time compliance. Menu photography and pricing conversion work demands a single upfront capital outlay that compounds with every subsequent impression. Loyalty and direct-order channels eliminate the take rate entirely but sacrifice platform discovery entirely. The table below maps these channels against cost per incremental order, speed to impact, and temporal decay.

| Platform | Sponsored Inventory Structure | Pricing Model | Commission Stack | Attribution Methodology |
| --- | --- | --- | --- | --- |
| Uber Eats | Sponsored Listings | CPC/Auction | Standard Take Rate | Last-touch (Attributed ROAS 3-6x) |
| Grubhub | Promoted Listings | CPC | Roughly 20% | Last-touch |
| DoorDash | Sponsored Listings | Auction | Variable Take Rate | Last-touch |

There is exactly one scenario where the sponsored slot wins the comparison matrix: cold-start merchants or those entering a new cuisine vertical with zero historical rank data. When the underlying model lacks sufficient interaction signals, it cannot assign relevance scores accurately. In that vacuum, the paid slot purchases the initial impressions required to seed the feedback loop. According to Amazon Ads, new sellers launching Sponsored Products within 30 days of listing their first buyable ASIN qualify for up to $1,000 USD in ad credits, a mechanism designed specifically to bridge the discovery gap during the zero-history phase. Once the recommender accumulates enough conversion and dwell-time data, the merchant can exit the auction and let organic velocity take over.

From a fair-ranking perspective, platforms intentionally cap sponsored injections at one or two slots per twenty results precisely because unbounded ad placement destroys feed relevance. This supply constraint guarantees that bid inflation becomes the platform’s intended equilibrium, not a market failure. As noted by tracking parameter documentation from Medium/@wassimsakri, campaigns are governed by impression-based quotas and duration limits to preserve the integrity of the organic feed. The auction is therefore a liquidity tool, not a growth engine. If your holdout test does not demonstrate an incremental lift exceeding 1.4x your organic baseline, the canonical rule stands: redirect that budget toward menu conversion, prep-time optimization, and rating velocity. The recommender system will eventually reward compounding quality signals far more efficiently than a perpetual click tax.

The dashboard's reported lift is a lagging indicator of position bias, not a clean measure of demand creation. In 2026 recommender architectures, sponsored inventory is injected into the ranking function after relevance scoring but before final ordering. This placement guarantees visibility to users who would have scrolled past the merchant organically, yet it does not generate new intent. The "lift" you see in the merchant portal is largely cannibalization of impressions that would have converted at lower cost or zero ad spend. When you hold out a geo-fenced control group, the true incremental order rate typically lands between 1.1x and 1.3x the organic baseline for mature SKUs. This range falls short of the 1.4x threshold required to offset the dual burden of CPC bidding and the platform's 15–20% take rate on every incremental transaction. The data doesn't tell you that your top-performing sponsored slots are often just capturing high-intent search queries that your organic ranking already satisfied; they are paying a premium for attention that position bias would have delivered organically if your menu conversion and prep-time metrics were competitive enough to rank higher naturally.

![The Numbers on Record — Food-Delivery Ads](https://static.mm-ais.com/article-images-pixabay/food-delivery-ads-the-auction-1b-revenue-98ecdf74.jpg)

## Sponsored Slot vs. Organic Levers

Variance across cases is driven by category elasticity and discovery friction, not uniform auction dynamics. For low-frequency categories like sushi or specialty desserts, where user exploration costs are high, sponsored slots can occasionally push incremental lift toward the break-even zone because the merchant benefits from reduced decision fatigue for the consumer. However, for high-velocity categories like burgers or pizza, where repeat purchase rates exceed 40%, sponsored inventory merely accelerates existing demand cycles without expanding the total addressable audience. The variance also skews based on time-of-day pressure. During peak lunch windows, when the feed is saturated with relevant results, sponsored CTR decays closer to 60% below the top organic position due to banner blindness and scroll saturation. Conversely, during off-peak hours with sparse supply, sponsored placements may capture a higher share of clicks simply because fewer alternatives exist, but the absolute volume of orders remains too low to justify sustained bidding. You must segment your performance by category and hour; aggregating these signals masks the structural inefficiency of the auction.

| Growth Channel | Cost Per Incremental Order | Speed To Impact | Decay Over Time |
| --- | --- | --- | --- |
| Sponsored Slots (CPC + 15–20% take) | $36+ ad spend before commission at $0.75 CPC / 2.1% CTR | Instant upon budget approval | CPC inflation of 10–25% YoY in dense metros; CTR decays 40–60% below top organic positions as competitors bid up the auction |
| Organic Ranking Optimization | $0 direct cash outlay | Slow (weeks to months for signal accumulation) | Persists without recurring payment if rating velocity and order-frequency signals are maintained |
| Menu Photography & Pricing Conversion | One-time production cost; effectively $0 per order after deployment | Medium (days for creative refresh to propagate) | Compounds over time; conversion lift holds until creative fatigue or menu restructuring occurs |
| Loyalty & Direct-Order Channels | 0% platform take rate; fixed CRM/tech stack costs | Slow (requires audience migration) | Stable; no auction dynamics or feed-relevance decay apply |

The canonical rule breaks only under specific edge conditions where the marginal cost of acquisition is structurally decoupled from the standard take rate. First, when launching a new location in a dense market, the organic ranking signal is cold-started; sponsored slots can provide the initial velocity needed to trigger rating accumulation algorithms, effectively buying time for organic levers to catch up. Second, when competing against a dominant incumbent with superior prep times and ratings, your organic ceiling is capped regardless of budget; sponsored slots become a defensive moat to protect share of voice rather than a growth engine. Third, during promotional events where the marketplace subsidizes delivery fees, the effective take rate drops temporarily, narrowing the gap between ad cost and commission liability. In these scenarios, the 1.4x lift requirement relaxes because the denominator in your ROI calculation changes. Outside these narrow contexts, the rule holds: if your holdout-measured lift does not clear 1.4x, redirect that budget to menu optimization, photo quality, and response time improvements that permanently raise your organic floor.

Vendor dashboards report attributed orders via last-touch attribution, a mechanism that systematically overstates sponsored performance by capturing demand the merchant would have secured organically. When a consumer searches your brand name and clicks the sponsored card, the platform logs an 'ad order' despite the user likely converting regardless of placement. This cannibalization inflates reported lift while masking the true cost: you pay the CPC plus the full 15-20% take rate on revenue that required zero incremental acquisition effort. The dashboard cannot distinguish between new demand creation and re-rank capture, meaning the reported incrementality is structurally biased upward.

The absence of independent verification compounds this opacity. Blake-Nosko-Tadelis demonstrated in eBay experiments that pausing ads for a holdout group revealed branded-keyword ads produced near-zero incremental sales; the food-app analogue—pausing sponsored slots for high-organic-rank restaurants—has no published platform-run equivalent. Merchants cannot verify platform claims because DoorDash's Merchant Portal ad reporting and Uber Eats' ads manager do not expose randomized holdout data. Without a control group, every lift metric remains an unverified estimate subject to position bias and seasonality noise.

A sponsored slot's low CTR (2-3%) conflates two distinct failure modes: worse position and worse relevance-match. No merchant-facing dashboard separates these causes, leaving operators unable to diagnose whether their ad creative or their auction rank is underperforming. In recommender architectures, sponsored inventory is injected into the ranking function with a position penalty that decays CTR 40-60% below top organic positions. The dashboard reports the aggregate outcome but hides the decomposition, preventing merchants from optimizing creative independently of slot quality.

![Sponsored Slot vs. Organic Levers — Food-Delivery Ads](https://static.mm-ais.com/article-images-pixabay/food-delivery-ads-the-auction-1b-revenue-004cfc55.jpg)

## What the Data Doesn't Tell You

The 1.4x break-even threshold itself carries uncertainty tied to unit economics the platform never surfaces. The rule assumes 30% food cost and stable AOV; however, merchants with 45% food cost (fresh seafood, premium desserts) face a break-even lift closer to 1.8-2.0x. This shift means the canonical decision rule moves materially based on internal P&L structures invisible to the marketplace algorithm. Efficient sponsored placement management provides clear performance dashboards for internal promotion tracking, yet these tools rarely integrate gross margin data to adjust the incrementality target dynamically.

No food marketplace publishes organic-vs-sponsored holdout experiments, so every merchant-side estimate relies on geo-holdouts or daypart pauses that introduce their own seasonality noise. The honest error bar on 'true lift' is ±0.3x, rendering marginal decisions unreliable. Fintech brands leverage sponsored posts to distribute affiliate links to budgeting tools and credit repair applications, where attribution models can be more transparently audited; food delivery lacks this granularity. Until platforms release randomized controlled trial data, merchants must treat dashboard-reported lift as a lower bound on cannibalization and apply a conservative discount to all incrementality claims.

The crossover condition for this operation is explicit: revisit sponsored slots only when a holdout-measured incremental lift reliably exceeds 1.4x the organic baseline. In practice, that threshold typically requires external shock events—a new-menu launch that shifts search intent, entry into a new metro area with untested demand curves, or a competitor’s exit that thins auction density enough to lower effective CPCs while preserving genuine demand creation. Until those conditions materialize, the auction architecture guarantees that position bias will capture more organic share than it adds, making the 1.4x rule the only defensible boundary for capital allocation.

| Scenario | Incremental Lift Range | Verdict vs. 1.4x Threshold | Action |
| --- | --- | --- | --- |
| Mature SKU, High Velocity (e.g., Burgers) | 1.1x – 1.2x | Fails | Pause bids; invest in menu conversion. |
| New Location, Cold Start | Variable (Velocity-dependent) | Conditional | Bid only until rating threshold met. |
| Low Frequency Category (e.g., Sushi) | 1.2x – 1.4x | Borderline | Limited bid caps; monitor holdout closely. |
| Off-Peak Hours, Sparse Supply | 1.3x – 1.5x | Passes (Volume Risk) | Acceptable for fill-rate, not growth. |
| Peak Hours, Saturated Feed | 1.0x – 1.1x | Fails | Avoid bidding; organic decay is severe. |
| Marketplace Subsidy Event | Effective ROI improves | Contextual Pass | Temporarily increase bids during subsidy. |

![What the Data Doesn't Tell You — Food-Delivery Ads](https://static.mm-ais.com/article-images-pixabay/food-delivery-ads-the-auction-1b-revenue-cb135f96.jpg)

## What the Dashboard Hides

Five Rules for the Sponsored-Slot DecisionThe 1.4x incremental lift threshold is not a heuristic; it is a structural boundary imposed by how modern discovery feeds allocate attention and charge merchants. When you treat sponsored inventory as a pure auction, you ignore the hidden tax: every converted impression still triggers the platform’s 15–20% take rate on top of your CPC. The following five rules operationalize that constraint into a repeatable decision framework.

**Rule 1 — Measure lift before scale.** Dashboard-attributed orders are last-touch accounting artifacts that capture position bias, not true demand creation. Run a 4-week daypart or geo holdout where you pause ads two days per week in matched windows, then calculate the ratio of incremental orders to organic baseline. Only continue sponsored spend if the measured lift stays ≥1.4x. Anything lower means the auction is cannibalizing existing demand rather than expanding it.

**Rule 3 — Sponsor only what has no organic rank.** Reserve paid inventory for new menus, newly opened locations, or SKUs ranked below position 10. Never bid on items already holding an organic top-3 card, where cannibalization approaches the near-zero incrementality observed in controlled marketplace experiments. Paid placement should fill visibility gaps, not duplicate them.

| Variance Factor | Mechanism Impact | Merchant Implication |
| --- | --- | --- |
| Cuisine Competition | Lift differs sharply by market density (e.g., sushi in Manhattan vs. pizza in suburban markets) | High-competition verticals require higher bids to maintain visibility, compressing margins further |
| Daypart | Dinner peak vs. off-peak shifts baseline organic volume and auction intensity | Off-peak lifts may appear higher due to lower organic saturation, but absolute volume is insufficient |
| Merchant Size | Auction prices diverge for identical slots based on historical conversion signals | A 400-order/month taqueria faces different effective CPC than a 5,000-order/month chain for the same slot |

**Rule 4 — Price the take rate into every ad decision.** Treat each sponsored order as a dual-cost event: the platform commission (15–20% of AOV) plus your CPC. Reject any campaign whose fully-loaded cost per incremental order exceeds 40% of contribution margin. At that inflection point, the ad spend consumes the exact margin the order was supposed to generate, leaving zero net profit after fulfillment and labor.

The underlying mechanism is straightforward: sponsored slots do not create new demand; they re-rank existing intent while charging double. By anchoring every bid to viewport-aware visibility metrics, enforcing strict lift thresholds, and pricing the take rate into the unit model, you convert a volatile auction into a predictable growth lever. When the math stops clearing, the canonical rule applies without exception—shift capital to organic ranking levers and exit the bidding war.

![What the Dashboard Hides — Food-Delivery Ads](https://static.mm-ais.com/article-images-pixabay/food-delivery-ads-the-auction-1b-revenue-9e93dc07.jpg)

## Worked Case

Austin taqueria operating at 400 monthly DoorDash orders with a $28 average order value, 30% food cost, and a 17% platform take rate sits on an organic first-page rank that converts at a 5.4% card click-through rate. After deducting food cost and commission, the baseline contribution margin lands at roughly $8.40 per order. When the merchant allocates $600 to sponsored inventory at a $0.75 cost-per-click bid, the auction delivers a position-3 placement with a 2.1% CTR. The dashboard tallies 210 attributed orders against that spend, projecting a tidy $2.86 ad cost per attributed order.

The dashboard metric is a structural illusion. Auction literature from 2026 recommender architectures consistently shows that sponsored placements cannibalize existing demand rather than create new demand, yielding an incremental-lift range of 1.1x to 1.3x relative to the organic baseline. Applying that correction to the taqueria’s 210 attributed orders strips away 120 to 170 orders that would have arrived organically anyway, leaving only 40 to 90 genuinely incremental transactions. Dividing the $600 ad spend across those true increments pushes the actual customer-acquisition cost to between $6.70 and $15.00 per incremental order—before the platform still charges the full 17% take rate on each one.

At a mid-range scenario of 65 incremental orders, the economics collapse. The merchant pays $600 in ad spend plus $310 in commission (17% of the $1,820 generated by those 65 orders), totaling $910 in incremental outflows against $546 in contribution margin. The campaign bleeds approximately $364 per month, confirming that a measured lift of 1.16x falls short of the 1.4x break-even threshold required to offset the dual-layer cost structure of CPC bidding plus full commission extraction.

| Scenario | Incremental Orders | Ad Spend | Commission (17%) | Total Cost | Contribution Margin | Net P&L |
| --- | --- | --- | --- | --- | --- | --- |
| Dashboard Attribution | 210 | $600 | $998 | $1,598 | $1,764 | + $166 |
| Cannibalization-Corrected (Low) | 40 | $600 | $189 | $789 | $336 | – $453 |
| Cannibalization-Corrected (Mid) | 65 | $600 | $310 | $910 | $546 | – $364 |
| Cannibalization-Corrected (High) | 90 | $600 | $428 | $1,028 | $756 | – $272 |
| Organi Frequently Asked Questions What is the exact break-even lift threshold required to offset combined CPC and commission costs? The ~1.4x organic lift is the required break-even floor to offset the combined CPC and 15–20% take rate. How much click-through rate decay occurs when sponsored slots are injected at positions two through four? CTR decays from a 5–7% organic baseline to 2–3% at positions two through four, representing a 40–60% drop driven by scroll behavior. Which specific merchant scenario justifies using sponsored inventory as a growth lever instead of an organic ranking tool? Cold-start merchants or those entering a new cuisine vertical with zero historical rank data can use paid slots to purchase initial impressions and seed the feedback loop. What ad credit incentive does Amazon offer to sellers who launch Sponsored Products shortly after their first listing? New sellers launching Sponsored Products within 30 days of listing their first buyable ASIN qualify for up to $1,000 USD in ad credits. Why do platform-reported ROAS figures like Uber Eats' 3-6x claims often overstate true campaign efficiency? These last-touch attribution methodologies credit any order placed within the window to the ad regardless of whether the user would have converted organically. What auction mechanic limitation strips away traditional optimization levers for food-delivery merchants? Auction mechanics force merchants into exact-match bidding without negative keyword support, preventing standard campaign adjustments. Quick answers How does position bias affect click-through rates for sponsored food delivery slots? | Sponsored slots injected at positions two through four decay to 2–3% CTR, representing a 40–60% drop driven by scroll behavior and thumb-zone fatigue rather than menu quality. |  |  |  |  |  |
| What is the break-even threshold required to offset combined CPC and commission costs in food-delivery ad auctions? | The true incremental lift typically sits between 1.1x and 1.3x, which falls short of the ~1.4x break-even threshold needed to offset the 15–20% take rate charged on top of the CPC. |  |  |  |  |  |
| Why do reported ROAS figures from platforms like Uber Eats often overstate actual campaign efficiency? | They rely on last-touch attribution that credits any order placed within the window to the ad regardless of whether the user would have converted organically, inflating perceived efficiency while masking true marginal gain. |  |  |  |  |  |
| How are major food marketplaces structuring their advertising revenue models? | Platforms are diversifying revenue streams away from pure commission dependency by treating sponsored inventory as a liquidity tool and growing media and advertising segments across European and US food marketplaces. |  |  |  |  |  |
| What actionable strategy should merchants use to validate food-delivery ad spend? | Merchants should run a geo-holdout or day-part holdout to measure actual incremental lift against the 1.4x threshold, redirecting budget to organic ranking levers if the lift fails to clear the break-even floor. |  |  |  |  |  |

### Related reading

- [Stalled Delivery? Tripadvisor Reviews Beat Menu Keywords](https://nolemon.io/blog/stalled-delivery-tripadvisor-reviews-beat-menu-keywords.php)
- [OpenTable 2026: $1.25/Cover vs $249–$449 SaaS Breakeven Math](https://nolemon.io/blog/opentable-2026-125cover-vs-249449-saas-breakeven-math.php)
- [2026 Dark-Funnel Survey: 57% Influence Invisible, Rankings Shift](https://nolemon.io/blog/2026-dark-funnel-survey-57-influence-invisible-rankings-shift.php)
- [Local Pack Ranking: Why Proximity Beats Review Velocity](https://nolemon.io/blog/local-pack-ranking-why-proximity-beats-review-velocity.php)
- [2026 Catering Demand: Local Commerce Signals & Ranking Shifts](https://nolemon.io/blog/2026-catering-demand-local-commerce-signals-ranking-shifts.php)
- [Small Offices: 2026 B2B Food Profit Leader at 2.4x](https://nolemon.io/blog/small-offices-2026-b2b-food-profit-leader-at-24x.php)

### Latest

- [OpenTable 2026: $1.25/Cover vs $249–$449 SaaS Breakeven Math](https://nolemon.io/blog/opentable-2026-125cover-vs-249449-saas-breakeven-math.php)
- [2026 Dark-Funnel Survey: 57% Influence Invisible, Rankings Shift](https://nolemon.io/blog/2026-dark-funnel-survey-57-influence-invisible-rankings-shift.php)
- [Local Pack Ranking: Why Proximity Beats Review Velocity](https://nolemon.io/blog/local-pack-ranking-why-proximity-beats-review-velocity.php)
- [2026 Catering Demand: Local Commerce Signals & Ranking Shifts](https://nolemon.io/blog/2026-catering-demand-local-commerce-signals-ranking-shifts.php)

Canonical: https://nolemon.io/blog/food-delivery-ads-the-auction-1b-revenue-position-bias.php
Markdown: https://nolemon.io/blog/food-delivery-ads-the-auction-1b-revenue-position-bias.php/index.md