Which proxy type should you use for ad verification?
Residential for web placements, mobile for in-app and carrier-targeted campaigns, datacenter for nothing more than redirect status checks. Ad stacks classify the network an IP belongs to before they decide what to serve, so the proxy type sets what a check can see. A residential IPv4 exit is a consumer ISP address and receives what a household in that city receives. A mobile proxy is a carrier address and receives carrier-targeted and in-app inventory. A datacenter IP is a hosting address, which invalid-traffic filters drop and cloaked campaigns serve a clean page.
For the checks themselves, see the ad verification proxies page. TrueProxies sells residential IPv4 (GB-based with city, region and ASN targeting, and Unlimited with country targeting) and Datacenter IPv6. It does not sell mobile carrier IPs or static ISP addresses, so this article is also honest about where to buy them. The decision table comes first; the rest explains each column.
- Residential IPv4: desktop and mobile-web display, search results, landing pages, cloaking re-checks, competitor monitoring. City and ASN precision on GB-based plans.
- Mobile carrier IP: in-app ads, carrier-targeted campaigns, app-install campaigns. Country and carrier precision, highest cost per check.
- Datacenter: redirect status codes on targets where cloaking is not a risk. Filtered as general invalid traffic elsewhere.
- Static ISP: long logged-in dashboards in a handful of countries. Not a verification exit.
- VPN: a datacenter exit with a system-wide tunnel and no per-check exit record. Not for verification.
Decision table by ad channel
| Channel | What the ad stack keys on | Use | Why | TrueProxies |
|---|---|---|---|---|
| Desktop display (publisher sites) | IP location and network type, browser signals | Residential | Slots fill for consumer-ISP visitors; datacenter IPs see fewer or different ads | GB-based (city, ASN) or Unlimited (country) |
| Mobile-web display and landing pages | IP location and network type, mobile user agent | Residential with a mobile user agent | Most mobile-web inventory is served to residential Wi-Fi as well as carriers | GB-based or Unlimited |
| Search ads (Google, Bing) | IP location, language, device, account | Residential, one query per fresh session | Search results are personalized and rate-limited | GB-based (city) for local campaigns |
| In-app ads (SDK inventory) | Carrier and device identifiers | Mobile carrier IP plus a real device or emulator | Inventory is bought by carrier and device; residential Wi-Fi is a different audience | Not offered; pair with a mobile specialist |
| Carrier-targeted campaigns | Carrier ASN | Mobile carrier IP | The targeting is the carrier | Not offered |
| Connected TV and OTT | Device and household IP | Residential (household IP) with the device | Household IP is the targeting key; the device is the hard part | Residential IPv4 for the network layer only |
| Affiliate and tracking links | IP location | Residential, or datacenter where cloaking is not a risk | Geo-routing keys on country; cloaked offers key on network type | GB-based or Unlimited; see the affiliate guide |
| Cloaking re-checks | Network type | Residential and datacenter, same city, compared | The difference between the two responses is the finding | Residential IPv4 for the consumer side |
The pattern: the exit should be the network type the campaign was bought for. Where a campaign targets consumers on any connection, residential IPv4 is the match. Where it targets carriers or app inventory, only a carrier IP is. The affiliate link guide covers the last two rows in detail.
Detection and invalid-traffic filtering
Ad verification vendors and ad exchanges filter invalid traffic in two classes. General invalid traffic (GIVT) is identified with routine lists: known crawlers, declared bots and traffic from known data-center IP ranges. Sophisticated invalid traffic (SIVT) needs behavioral analysis. The Media Rating Council's invalid-traffic guidelines define both, and the practical consequence for a verifier is simple: a check from a datacenter IP is filtered as GIVT at many ad calls, so the page you receive is not the page a consumer receives. The glossary entry on GIVT and SIVT has the definitions.
Residential and mobile IPs are not classified as data-center traffic, which is why they exist as proxy products. They are not invisible. Ad stacks and fraud tools also score IP reputation, and a residential IP that many customers have reused for automation carries a history. Publishers read browser fingerprints, cookies and behavior. No proxy type makes a check identical to a person, and a provider who promises that is describing something else. What a residential or mobile exit does is remove the network-type difference, so that the remaining differences are ones you control: profile, locale, timing and repetition.
Cloaking is the special case. A cloaked campaign serves reviewers, verification vendors, offices and datacenters a compliant page and consumer IPs the real payload, and the detection method is the comparison itself: load the same ad and landing page from a datacenter exit and a residential exit in the same city and record both. The ad cloaking glossary entry has the check; Google's Cloak of Visibility paper has the measurements.
Geo precision
Residential IPv4 goes furthest. On TrueProxies GB-based plans the proxy username selects a country, a city, an ISO 3166-2 region or an ASN, so a campaign bought for Munich or for Deutsche Telekom subscribers can be checked from that exit; Unlimited plans select the country. The accepted values are listed in the connection reference, and a wrong value is refused rather than silently ignored, so a typo never bills a check from the wrong place.
Mobile carrier proxies select a country and a carrier; the carrier is the precision, and city selection is rarer because a carrier's addresses geolocate to the network, not the handset. Datacenter proxies select the country or city of the hosting facility, which for TrueProxies Datacenter IPv6 is one of ten cities chosen at checkout. Static ISP proxies come in a few countries at fixed addresses. Neither residential product on TrueProxies offers ZIP or postal-code targeting, and the location pages state what has been verified for each market with the date.
Cost per 1,000 checks
Data per check was measured on 30 September 2026 from a direct connection: 0.8 MB at the median for a redirect-chain check that reads whole documents, 3.8 MB for a full render with a screenshot, 7.6 MB for a YouTube watch page with eight seconds of playback. On TrueProxies GB-based traffic packs, 1 GB for $1.64 down to $0.60/GB on the 5,000 GB traffic pack, the residential column below follows from those medians. Mobile proxies are usually sold per port or per gigabyte at several times the residential per-GB rate; the column shows the shape, not a vendor's price, because those prices change and are not TrueProxies' to publish.
| Check type | Residential IPv4 GB-based, 1 GB traffic pack | Residential IPv4 GB-based, 5,000 GB traffic pack | Mobile carrier (typical market shape) | Datacenter |
|---|---|---|---|---|
| Redirect chain only | $1.31 | $0.48 | Several times the residential rate per GB, or a fixed price per port | Cheapest, but filtered or cloaked where it matters |
| Full render with screenshot | $6.23 | $2.28 | As above | As above |
| Video-heavy page | $12.46 | $4.56 | As above | As above |
For recurring country-level monitoring at volume, Residential IPv4 Unlimited moves up to 4.5 GB an hour at 10 Mbps for $4.08 an hour or $7.35 a day, which caps at about 28,000 full renders a day at full utilization. The cost per check article has the method, the 90th percentiles and the break-even against traffic packs.
Where TrueProxies fits, and where to buy mobile instead
TrueProxies fits the residential rows: web display, search, landing pages, cloaking re-checks, competitor monitoring and affiliate audits, with city, region and ASN targeting on GB-based plans and country targeting on Unlimited. Every check comes with the evidence templates and field dictionary on the ad verification proxies page, so results from different markets stay comparable. Datacenter IPv6 is sold for other workloads; for ad checks it only suits redirect status codes on dual-stack targets where cloaking is not a risk, and it cannot load hosts without AAAA records at all.
For in-app and carrier-targeted campaigns, buy a mobile carrier proxy from a mobile specialist and pair it with a real device or an emulator that carries the app's identifiers; several residential providers also resell carrier pools with per-carrier selection. Judge any of them on the same batch you would run here: successful responses, challenge rate, exit accuracy against the carrier you asked for, and cost per check. Keep the two evidence sets in the same template so a campaign report can show the web placement and the in-app placement side by side.
FAQ
Do I need mobile proxies for ad verification? Only for in-app inventory and carrier-targeted campaigns. Mobile-web placements and landing pages are served to residential connections too, and a residential IPv4 exit with a mobile user agent checks them.
Are datacenter proxies good for ad verification? Rarely. Known data-center ranges are filtered as general invalid traffic and cloaked campaigns serve them the clean page. They are fine for checking redirect status codes on targets where neither applies.
Can I use a residential IP VPN for ad verification? A VPN whose exits are residential removes the network-type problem but still tunnels the whole device, offers country-level choice at best, and records nothing per check. A residential proxy in a script selects the exit per check, targets a city or ASN, and logs the exit with the evidence.
Rotating or sticky sessions for ad checks? Rotating for independent single-page checks, so each sample comes from a different exit. Sticky for a landing page followed by a checkout or sign-up step, so the flow stays on one exit; set the session lifetime to the length of the flow. Rotating proxies vs sticky sessions explains both.