mediabuyer
Back to news

Economics

How to actually read a Tonic RSOC payout report

An operational guide to the Tonic RSOC reporting interface — what each column means, where the fraud signals hide, and how to action the data instead of just staring at it.

By Eyal RosenthalMay 6, 202614 min readAI-assisted research

A new search-arbitrage operator opens the Tonic dashboard for the first time and is greeted by a dozen columns of numbers, half of them blank, and the helpful realization that nobody has ever explained any of them. Tonic's help center covers the basics but skips the operational interpretation — the part where you actually figure out what to do next. This piece is the missing manual. I am writing it after spending more time than I care to admit staring at these reports while running test campaigns and pulling apart the numbers with operators who run more spend than I do.

Quick orientation: what Tonic is and isn't

Tonic.com is the publisher-and-arbitrage interface for System1, the publicly-traded search-monetization company (NYSE: SST). Their full corporate disclosure lives in their SEC filings, which are worth reading for the segment-level economics of the RSOC business model. Tonic itself is the operational front-end where you create "campaigns" (really: search feeds tied to topic categories), get tracking links, set up domains, and pull payout reports.

The product evolved from System1's older RSOC (Revenue Share on Click) model and the affiliate-search-arbitrage industry that grew up around it from roughly 2017 onward. The reporting interface in 2026 reflects that history: it is dense, jargon-heavy, and assumes you know the model. If you do not, the dashboard will not teach you.

The columns, in plain English

A typical Tonic RSOC report has the following columns. I'll go through what each one means and where the operational nuance lives. Names and exact column ordering may vary depending on which Tonic interface revision you are looking at and which campaign type (TDFR vs RSOC vs hybrid) — the Tonic help docs describe the variants. The general logic is the same across versions.

Campaign and Tracking ID / subID

A "campaign" in Tonic is a configured search feed — typically tied to a topic category (auto, insurance, loans, recipes, news) and a specific search partner backend (System1 owned-and-operated, Yahoo Partner Network, or a custom partner). The tracking ID and subID fields are the URL parameters you append when you send traffic — ?subid=outbrain_us_creative42 style — and they are the only way you'll know which traffic source produced what. Use them religiously and at high resolution. A typical disciplined operator passes through five to seven dimensions in subIDs (network, country, device, ad-set, creative, lander variant, day) so the post-hoc analysis works.

Sessions (sometimes Visits or Inbound Clicks)

This is the number of click-throughs from the traffic source that landed on your pre-lander or directly on the search feed. Tonic measures it via a 1x1 pixel or via redirect logging, depending on how the integration was set up. The number you see on Tonic should be roughly equal to (but slightly less than) the number of clicks the traffic source charged you for. If Tonic sessions are dramatically lower (more than ~10-15% lower) than Outbrain or Taboola clicks, you are leaking traffic — usually slow page load, broken redirects, or aggressive ad-blocker behavior. This is the first place to look when revenue mysteriously drops.

Clicks (the search-feed clicks, sometimes Monetized Clicks)

This is the count of users who reached the search feed and clicked an ad-blended search result. This is the metric that actually monetizes. The ratio of Clicks to Sessions is your search-feed CTR — typically 50-85% in real campaigns. If yours is below 50%, you have either a broken pre-lander integration, a low-intent traffic source, or a search-feed query that doesn't resonate.

RPM or EPC (revenue per thousand sessions, or earnings per click)

The big-picture monetization metric. Tonic reports this both ways depending on the view. RPM is revenue per thousand sessions — useful for comparing to a traffic-source CPM. EPC is earnings per click — useful for comparing to traffic-source CPC. Watch which one is being shown; mixing them up is the most common newbie mistake.

For tier-1 (US, UK, CA, AU) traffic in monetizable verticals, healthy 2026 EPC ranges I've heard cited are roughly $0.18-$0.55 per inbound click. Below that range you have a quality, vertical, or geo problem. Above it, you either have a great account or a vertical I should be running.

Revenue (gross vs net)

This is the dollar figure. Critical: confirm whether the revenue figure shown is gross (before Tonic's revenue share) or net (your share, after Tonic's cut). Tonic's standard rev-share is in the 20-30% range — meaning you keep 70-80% of gross — but the exact percentage is per-account and not always visible in the UI. Pull your contract or ask your account manager. Comparing gross revenue from one campaign to net revenue from another is how operators talk past each other for weeks.

Pending vs Approved vs Reversed

This is where it gets operational. Tonic, like all RSOC platforms, holds a portion of revenue in "pending" until the search partner confirms the click was high-quality. Search partners can and do reverse clicks they later determine were low-quality, fraudulent, or non-monetizing. The reversal rate is one of the single most important health signals on your account.

A clean tier-1 campaign has reversal rates in the 1-5% range. Anything 8%+ is a problem. Anything 15%+ and you are about to be paused or have your payout rate cut. Reversals show up as line items in the report — you have to look. They are not always loud.

Country / Geo

The geo breakdown by country, sometimes by region. Critical for catching cost-vs-revenue mismatches: a campaign that's bidding "US/CA" on the traffic source but somehow showing 30% of sessions from Mexico or India is a campaign with a leak. The leak is usually misconfigured geo on the native network or a publisher who is mis-reporting traffic origin. Either way, that 30% is going to monetize at one-tenth the rate of the tier-1 sessions, and your blended EPC will look terrible.

Device / Platform

iOS, Android, desktop. Same diagnostic value. If you bid mobile-only on Outbrain and 20% of your Tonic sessions are showing as desktop, somebody is misclassifying traffic. Usually it's a matter of how user-agent parsing is happening between the click-redirect chain and Tonic's pixel — but it can also indicate bot or fraud traffic.

Source / Referrer

The HTTP referer chain. In a clean integration, this should show the traffic source's redirect domain or the publisher domain. If you're seeing direct (no referer) for a meaningful percentage of clicks, you have either an ad-blocker problem, a same-origin-policy issue, or — and this is the painful one — bot traffic that's stripping referers.

The four fraud signals that show up in the report

Search-feed monetization is sensitive to fraud and the partners actively scrub it. The patterns to watch for in your own report:

1. Low search-feed CTR with high session count. Bots hit the lander at speed and don't click. If your sessions number is climbing but your monetized-click number is flat, you are getting non-human traffic. Cross-reference with the publisher subIDs on the traffic source — often it's one or two specific publishers. Block them.

2. Geo mismatch. Sessions from a country you're not bidding on. Either the publisher is mis-reporting (block them) or the redirect is leaking through a proxy/CDN edge in another region (fix the integration).

3. Device mismatch beyond a few percent. Same logic. iOS bidder, but 25% Android sessions = something is wrong.

4. Reversal-rate creep. Reversals were 2% last week, 4% this week, 8% next week. The search partner has detected something. You have one to three weeks before the payout rate gets cut. Find the source — usually a single bad publisher subID or a creative that's attracting bot traffic — and kill it.

These four are not all of fraud, but they are the four you can catch in a normal Tonic report without specialized tooling. The System1 quality and compliance documentation describes additional fraud-detection layers they apply server-side.

A synthetic example, week-over-week

Let me walk through a synthetic but representative report comparison. (Real numbers are not mine to publish.)

| Week | Sessions | Search-feed clicks | Pre-lander CTR | Gross revenue | Net revenue | EPC (net) | Reversal % | |---|---|---|---|---|---|---|---| | Week 1 | 142,000 | 95,000 | 67% | $24,000 | $17,000 | $0.12 | 2.1% | | Week 2 | 156,000 | 99,000 | 63% | $25,500 | $18,000 | $0.115 | 2.4% | | Week 3 | 168,000 | 99,000 | 59% | $26,000 | $18,200 | $0.108 | 4.1% | | Week 4 | 175,000 | 95,000 | 54% | $24,500 | $17,000 | $0.097 | 6.8% |

What is happening here is a slow-motion train wreck. Sessions are growing — the operator probably scaled spend. Pre-lander CTR is dropping — meaning incremental sessions are lower-intent. Reversal rate is climbing — meaning the search partner is increasingly scrubbing the new traffic. Net EPC is dropping in proportion. By week 4, the marginal session is monetizing at maybe $0.06, and the operator is probably still buying it at $0.13.

The intervention is at week 2 or week 3, not week 4. The signal is in the report. Most operators do not look at the report at this resolution.

What to action on

A weekly Tonic-report review should output, at minimum:

  1. Network and creative-level kill list. Subs whose net EPC is below your floor for two consecutive weeks. Kill them on the traffic source — do not wait.
  2. Network and creative-level scale list. Subs whose net EPC is above your target with stable reversal rates. Scale spend on these immediately.
  3. Geo and device anomaly check. Anything that looks structurally off vs your bid configuration on the native network. Reconcile.
  4. Reversal-rate trajectory. Plot reversal rate by week for each campaign. Anything trending up is a near-term ban risk.
  5. Account-manager outreach. If your aggregate EPC has dropped 15%+ over four weeks with no obvious cause, ping your Tonic account manager. They will sometimes confirm whether the search-partner backend has rebalanced rates (it does, periodically) or whether something on your side has triggered a quality-score reduction.

The conversation with your Tonic AM

Account managers at Tonic, like at System1's other partners, have visibility into things you do not. They can see (anonymized) whether your account's EPC is in the top quartile or bottom quartile vs your peer set. They can see (with permission) which verticals you've been throttled on. They can sometimes push for a custom rev-share if you're scaled enough.

Things to ask, that they will sometimes answer:

  • "Is my reversal rate higher than peer median?" (Yes/no answer.)
  • "Has the search partner I'm pointing at had a recent payout-rate change?" (Often yes.)
  • "Can you push for a 2-3% rev-share improvement on my main vertical?" (At $50K+/month in revenue, sometimes yes.)
  • "Is there a non-throttled vertical I'm not running where my profile would qualify?" (Sometimes yes.)

What they will not answer: the exact backend partner, the exact gross-CPC-per-keyword, the exact rev-share negotiated with the search partner. NDAs.

What I get wrong about Tonic, often

A few things I have personally been wrong about that have cost me time:

Confusing "Pending" with "Frozen." Pending revenue is normal — it just means the click hasn't been confirmed yet. Frozen revenue is what happens when an entire campaign or account is under compliance review. They look similar in the UI but they're operationally different.

Reading EPC at low session counts. A campaign with 200 sessions and a $0.40 EPC is not a $0.40 EPC campaign. It's a sample of 200 with a confidence interval that includes $0.10 and $0.70. Wait for at least 5,000 sessions, ideally 20,000, before treating EPC as a real number.

Ignoring the time-zone delta. Tonic reports in PST. Your traffic source might report in UTC, in your local time, or in account-creator local time. Daily-level comparisons across systems require timezone normalization or you get phantom 8-hour gaps.

Not breaking out by hour-of-day. Aggregate EPC hides huge intraday swings. A campaign that's a great average across 24 hours can be a disaster from 10 PM-6 AM and a goldmine from 9 AM-5 PM. Dayparting on the traffic source is the fix.

How Tonic fits in the broader RSOC ecosystem

A bit of context on where Tonic sits relative to peers helps with the operational reading. The RSOC space has roughly four buckets of partners:

Direct-from-search-engine programs. Microsoft and Yahoo (the two that still run partner-network programs at scale; Google has substantially curtailed AdSense for Search distribution since the late 2010s) operate their own publisher partnership tracks. The Microsoft Audience Solutions documentation describes the Microsoft side; Yahoo's publisher solutions covers Yahoo. Direct relationships pay better revenue shares but require larger volume and more direct quality oversight.

Aggregator partners (System1/Tonic, Bodis, Domain Holdings, Sedo, AdMarketplace). These companies hold direct contracts with the search engines and resell access to publishers and arbitrage operators. They take a margin (typically 20-30%) but provide tooling, faster onboarding, and a single interface. Tonic is the most-used of these in affiliate search-arbitrage; the others address adjacent niches (Bodis is more domain-monetization-focused; AdMarketplace is more enterprise).

Native-platform-integrated search. Some native networks have begun offering RSOC-equivalent products integrated into their bidder. These have been quietly shipping over the last 18 months and are worth tracking but are not yet meaningful share.

Affiliate-style "search results" partners. A long tail of smaller intermediaries with worse rates and worse compliance posture. Generally avoid.

For an operator, the practical implication is that the Tonic dashboard is one of typically two-to-four reporting interfaces you'll be looking at simultaneously. Reconciling them across systems — same-period revenue, click counts, reversal patterns — is the operational work.

Reporting hygiene checklist

A few habits that distinguish operators who run this profitably from operators who don't:

  1. Pull a fresh report at the same time every day. Tonic's data backfills somewhat for the first 24-48 hours after a session, so today's number isn't fully resolved. Comparing yesterday at 10 AM to today at 4 PM gives you false signal. Same time, every day.
  1. Reconcile to traffic-source numbers within 1-2 days. Your Outbrain/Taboola click count vs. your Tonic session count should be roughly aligned (with the click-loss factor described earlier). A growing gap is operational signal.
  1. Separate test vs scale spend in the data. If you're spending $100/day testing a new creative inside a campaign that's spending $2,000/day on existing creative, the blended numbers are useless for either decision. Sub-ID resolution is non-optional.
  1. Track gross and net revenue separately, always. Confusing them is the most common mistake in cross-operator conversation.
  1. Track reversal rate as a leading indicator of payout-rate cuts. Reversal trend is more diagnostic than current EPC. By the time EPC drops, you're already in trouble. The reversal trend gives you a 2-4 week head start.
  1. Save raw exports, not just dashboard screenshots. When you need to investigate a problem six weeks later — "why did our EPC drop in week of March 14?" — you need the underlying data, not a screenshot of an aggregate. Set up an automated export to S3 or Google Drive on a daily schedule.
  1. Build a one-page weekly P&L that pulls from all sources. Outbrain spend, Taboola spend, Tonic gross, Tonic net, hosting/tooling cost, net margin. One page, every Monday morning, before you make any optimization decisions for the week. The format is less important than the discipline of doing it.

What this looks like as a system

Mature search-arb operations end up building a small ETL pipeline: traffic-source API exports → Tonic API exports → tracker exports → reconciliation database → daily/weekly summary dashboards. The cost to build this in 2026 is dramatically lower than it was three years ago — between off-the-shelf tools (Voluum, RedTrack), simple Python or Node scripts, and modern data warehouses (BigQuery, Snowflake at smaller scale, even SQLite for solo operators), a single capable engineer can build the full stack in a couple of weeks.

The operators who don't have this stack are working blind. The ones who do typically catch issues two-to-three weeks earlier than the ones who don't, and the difference compounds. Tonic's reporting alone is necessary but not sufficient; it's one input into a multi-source picture.

If I were starting a search-arb operation today and could only do one infrastructure thing first, it would be the reporting reconciliation pipeline. Everything else — creative production, lander optimization, account warm-up — is downstream of having clean data on what's actually happening.

A note on Tonic's product evolution over the past two years

The Tonic interface in 2026 is meaningfully different from the Tonic interface in 2023. A few notable shifts:

More backend partner choices. Tonic now offers multiple search-partner backends in a single account interface. Operators can route different campaigns to different partners (System1 owned-and-operated, Yahoo Partner Network, others) and compare performance side by side. Two years ago this required separate accounts.

More compliance pre-screening. Tonic now applies its own compliance review before traffic hits a search-partner backend, which has reduced the rate of post-click clawbacks but has also lengthened the new-account onboarding cycle.

Better API access. The Tonic API documentation has expanded meaningfully. Programmatic pulls of session, click, revenue, and reversal data are now possible, where previously some of these required dashboard scraping.

Improved cohort-level reporting. Day-and-week cohort views give better visibility into how a given week of traffic is monetizing relative to peer weeks. This is the right tooling for catching slow-motion problems like the one in the synthetic example earlier.

The implication for operators familiar with the older Tonic interface: it's worth re-reading the help docs in 2026, because there are features that didn't exist 18 months ago that meaningfully change the operational workflow.

What I'd build if I had to redesign the Tonic dashboard

A self-indulgent closing note. The Tonic dashboard, as a piece of operator UX, has improved over time but still leaves work on the table. If I had to redesign it, the priorities would be:

  1. A single-glance health indicator per campaign. Trend on session volume, EPC, reversal rate, and pre-lander CTR — all on one widget per campaign. The current dashboard makes you click through.
  2. Automatic flagging of statistical anomalies. Reversal rate up 50% week over week? Auto-flag. Pre-lander CTR down 10% on a consistent traffic source? Auto-flag.
  3. Better defaults on date-range filters. "Last 7 days" should be the default, not "Today" — which encourages low-N decisions.
  4. First-class support for sub-ID hierarchies. The current sub-ID handling is flat; multi-level hierarchies would unlock cleaner attribution.
  5. Cross-campaign comparison views. Comparing two campaigns side-by-side across the same dimensions is currently awkward.

This is operator-side wishlist, not a public roadmap. But it's the kind of feedback Tonic's product team probably hears regularly, and a sense of which features would land best is helpful for any operator running across multiple search-partner platforms.

Further reading and primary sources

Editor's note: AI-assisted research; written and reviewed by Eyal Rosenthal. Sources cited above. Send corrections to corrections@mediabuyer.site.