Category: Marketing

IP targeting is a form of advertising that matches physical addresses, like the ones on your customer list or direct mail file, to the internet connections at those locations, then serves display or video ads to devices on those networks. It still works in 2026, but not the way most vendors describe it. It reaches households and businesses, not specific individuals, and its accuracy depends heavily on the provider and the data behind the match.

That distinction matters more than it used to. We published the first version of this article back in 2017, when IP targeting was routinely pitched as “true one-to-one marketing.” The industry has since produced independent research that tells a more complicated story, and if you are considering spending money on this tactic, you deserve the complicated story.

Here is what IP targeting actually is today, where it genuinely performs, where it falls apart, and how to test it without wasting your budget.

“IP Targeting” Now Means Five Different Things

One reason this topic confuses business owners is that marketers use one term for five different practices. Before you buy anything, know which one you are actually being sold.

1. IP geolocation. An ad platform uses an IP address to estimate a device’s general location, like a state or city. Google Ads uses IP addresses this way for geographic targeting, and Google itself notes that an internet provider may assign the same IP to many devices. This is broad targeting, not address-level targeting.

2. Address-to-IP household activation. You provide a list of physical addresses. A specialized provider such as El Toro attempts to match those addresses to residential IPs and serves ads to matched households. This is what our original article described, and it is what most local businesses mean when they say “IP targeting.”

3. Household and cross-device extension. A provider links multiple devices (the TV, the laptop, the phones) to one household using the IP as one signal among several. Identity platforms like FreeWheel are explicit that IPs should support this process, not carry it alone.

4. IP-based attribution. An ad is shown to one device and a conversion happens on another device using the same home connection, so the platform credits the household. Microsoft offers this and openly documents its limits: it works best when the IP is stable, unique to one household, and not masked by a VPN.

5. Reverse-IP business identification. Tools identify which company an anonymous website visitor likely works for based on the office network they browsed from. Useful for B2B sales intelligence, but it is a different workflow from ad delivery, and it identifies a company, not the person.

One practical note that surprises many owners: you generally cannot upload a list of IP addresses into a standard Google Ads campaign and target them directly. Google documents IP-based location inference and IP exclusions, not IP inclusion targeting. True address-to-IP activation requires a specialized provider or programmatic platform.

How Address-to-IP Targeting Works

The core workflow has not changed much since our original article, and it is still worth understanding:

1. You provide names and postal addresses: past customers, lapsed customers, a purchased or compiled prospect list, or your direct mail file.
2. The provider runs the list through a matching system that attempts to link each address to the IP address of the internet connection at that location.
3. Matched households become an audience, and your display or video ads are served to devices on those networks through programmatic ad inventory.
4. Reporting shows delivery, engagement, and (with some providers) household-level attribution.

Where the 2017 version of this article went wrong was in what happens at step 2. We described the match as essentially exact, down to the suite number. The evidence now available says otherwise.

How Accurate Is IP Targeting, Really?

This is the question the industry avoided for years, and in late 2025 it got a serious answer.

A study commissioned by the Coalition for Innovative Media Measurement and Go Addressable, conducted by the data-validation firm Truthset, benchmarked nearly one billion commercial IP linkage records from six major data providers against actual records from internet service providers. The findings were blunt:

• IP-to-postal-address linkages were accurate about 13 percent of the time on average
• IP-to-email linkages were accurate about 16 percent of the time
• Different data providers agreed with each other on a given postal linkage only 6.4 percent of the time
• Commercial files contained roughly three times more IPs per address than ISP records supported, a sign of systematic over-linking
• IPv6 addresses, a large share of modern traffic, were badly underrepresented

Now the necessary qualification, because this study gets misquoted in both directions. It does not prove that every IP targeting campaign is 13 percent accurate. It evaluated six commercial linkage datasets of the kind used in connected TV and audience measurement. A provider with direct ISP relationships or deterministic first-party data can perform much better. The report’s actual conclusion is that IPs should not be the primary or sole method of targeting and measurement.

What it does prove is that a vendor’s “95 percent match rate” claim tells you almost nothing by itself. A match rate is how many of your addresses the vendor linked to some IP. Accuracy is whether those links are correct. Those are very different numbers, and the gap between them is where budgets go to die.

Why matches fail

The failure modes are mundane and mechanical:

• Dynamic IPs. Your internet provider can reassign your home’s IP at any time. A router reset can break the link between the ad impression and the conversion.
• Shared networks. Apartment buildings, office suites, and public networks can put many households behind one IP. Microsoft’s attribution documentation states plainly that units sharing an IP cannot be told apart.
• VPNs and proxies. A VPN swaps in a different network’s address entirely. This is increasingly common in professional households.
• Cellular connections. Mobile carriers use large shared, rotating address pools, which is why serious attribution products filter cellular traffic out.
• Hybrid work. The employee you want to reach at “the office” may be on home Wi-Fi, a VPN, or a phone. This is the single biggest reason office-based B2B IP targeting is weaker than it was in 2017.

Can It Still Work? Yes, Under the Right Conditions

None of the above means the tactic is dead. It means the tactic is conditional.

The published results that hold up best all share the same shape: a known audience list, a coordinated offline and online message, and a control group. A Lawn Doctor campaign reported a 21 to 44 percent sales lift when IP-matched digital ads were layered onto direct mail. El Toro’s multi-campaign analysis reported targeted households were 33 percent more likely to purchase than mail-only controls. These figures come from vendors and agencies selling the service, so treat them as evidence the approach can work, not as a benchmark your campaign should expect. Methodologies differ, costs are not always fully disclosed, and attribution can assign credit without proving cause.

The honest summary: IP targeting can be effective, and the only way to know if it is effective for you is to test it against a holdout group on your own list.

Where IP Targeting Fits Best

Based on the current evidence and our own client work at DMG, these are the applications worth testing:

Reinforcing direct mail. Still the strongest use case, and the one we highlighted in 2017. You already know exactly which households get the mailer. Digital ads before, during, and after delivery create repeated exposure around one offer. Home services, automotive, nonprofits, and local memberships fit well. If you are already investing in PPC advertising, this layers on the same programmatic muscles with a tighter audience.

Reactivating past customers. A clean first-party customer list, minus opt-outs, with a genuine comeback offer. Household-level reach is fine here because the message works for anyone in the house.

High-consideration local services. Roofing, HVAC, remodeling, window replacement. High ticket values, defined service territories, long consideration windows that reward repeated exposure. These are the same home services businesses where we see multi-touch campaigns outperform single-channel spends.

Account-based B2B, selectively. Reaching a known company’s office network can still work, but hybrid work has weakened it. Modern B2B programs pair reverse-IP company identification with person-level outreach rather than relying on office IPs alone.

Where to Skip It

Save your money when:

• The campaign needs to reach a specific person, not a household
• Your list is heavy on apartments, shared buildings, or office suites
• Your audience is mobile-first or VPN-heavy
• The list is too small to measure results with any statistical confidence
• The campaign touches sensitive categories (health conditions, finances, religion) where household-level ad delivery creates privacy risk
• A cheaper method, like standard geographic targeting, contextual placement, or uploading a hashed email list as a platform custom audience, can hit the same goal
• The vendor cannot explain how matches are made and refreshed

That last one deserves emphasis. A provider who gets vague when you ask about methodology is answering your question.

IP Targeting vs. the Alternatives

The 2017 version of this article said IP targeting was “superior to geo-fencing.” That framing was wrong. They solve different problems, and the right choice depends on what data you are starting with.

The Privacy Part You Cannot Skip

This section did not exist in our original article. In 2026 it has to.

IP addresses can be personal data. California’s privacy law explicitly lists IP addresses among identifiers that may qualify as personal information when linkable to a person or household.

New Jersey businesses have a state law to think about. The New Jersey Data Privacy Act took effect in 2025 and gives qualifying consumers the right to opt out of processing for targeted advertising, with required support for universal opt-out mechanisms. Whether a specific IP campaign falls under specific provisions depends on your business size, data sources, and exemptions. This is general information, not legal advice, and campaigns in sensitive categories deserve an actual legal review.

“Cookieless” is not a privacy pass. Vendors love pitching IP targeting as the cookieless answer. It can operate without third-party browser cookies, true. It still uses data about identifiable households, and consent, disclosure, and opt-out obligations still apply.

Sensitive locations get heightened scrutiny. Federal Trade Commission enforcement has targeted the use of location data around healthcare facilities, places of worship, and similar sites. Those cases mostly involved mobile location data rather than household IP matching, but the direction of regulatory attention is clear.

The DMG IP Targeting Fit Test

Before we recommend this tactic to any client, we run the campaign idea through ten questions. Score one point for each yes.

1. Is the audience list first-party or lawfully sourced, with opt-outs honored?
2. Is the goal household-level (an offer for the home or business), not person-level?
3. Are the postal addresses current and standardized?
4. Will the vendor explain their matching methodology in writing?
5. Is any of the match data deterministic (ISP-verified or authenticated) rather than purely inferred?
6. Is the list large enough to produce measurable results?
7. Is there a trackable conversion (calls, forms, sales) tied to the campaign?
8. Can you hold out a control group and compare?
9. Is the category non-sensitive?
10. Is the total cost competitive with the simplest alternative that could reach the same audience?

8 to 10: worth a structured pilot. 5 to 7: fix the weak points first. Below 5: spend the budget elsewhere.

Six Questions to Ask Any IP Targeting Vendor

If the pilot makes sense, these questions separate serious providers from resellers of the commercial linkage files the CIMM study evaluated:

1. What exactly does your accuracy percentage measure, and against what source of truth?
2. Is your address-to-IP data deterministic or probabilistic, and how recently was each match verified?
3. How do you handle dynamic IPs, VPNs, cellular traffic, and multi-unit buildings?
4. What match rate should I expect for a list like mine, and what happens to unmatched records?
5. Does your reporting separate attributed conversions from incremental lift, and can you support a true holdout group?
6. Are data costs, media costs, and platform fees broken out separately?

How to Run a Test That Actually Proves Something

1. Define the business question. Not “can we target these addresses” but “does adding household digital ads produce more qualified conversions than the campaign we already run?”
2. Clean the list. Remove duplicates, invalid addresses, opted-out customers, and anyone outside the service area.
3. Get the vendor’s match report before launch. Uploaded records, matched records, methodology, estimated reachable households.
4. Hold out a control. Randomly withhold a statistically meaningful slice of the list from the digital layer. Mail-only versus mail-plus-digital is the classic design.
5. Coordinate the creative. Same offer, same look, same landing page across mail and digital. The reinforcement effect is the whole point.
6. Measure business outcomes. Incremental leads, incremental sales, and cost per incremental customer. Clicks and impressions diagnose creative; they do not prove impact.

If a campaign cannot support this structure, that is a signal the list is too small or the tactic is wrong for the goal.

Score Your Campaign Before You Spend

DMG will review your audience list, campaign goal, and current marketing, and give you an honest assessment of whether IP targeting deserves a pilot or whether something simpler will get you further. No pitch deck, no pressure.

Request a Free Strategy Consultation →

The Bottom Line: Treat It as a Test, Not a Promise

IP targeting earned its place in the toolkit, and it keeps that place in 2026 for one specific job: putting coordinated digital ads in front of households you already know you want to reach, especially alongside direct mail. What it lost is the mythology. It is not one-to-one marketing, it is not automatically superior to anything, and no vendor claim substitutes for a controlled test on your own list.

If you are weighing IP targeting against other ways to spend the same budget, we can help you run the numbers honestly. DMG will review your list, your goal, and your current campaigns, and tell you whether this tactic deserves a pilot or whether something simpler will get you further. Request a free strategy consultation and we will give you a straight answer either way.