TapCub is live — analytics, insights and live chat, free on one platform
Glossary

What is a demand-side platform (DSP)? Definition and how to check its reports

A demand-side platform is software advertisers use to buy impressions from many sources. Where it sits among ad-tech terms, and why its conversion numbers need reconciling with your analytics.

Oliver Reed Reviewed by Emma Collins Also called DSP Last updated Nov 19, 2026 Glossary

Quick answer

A demand-side platform (DSP) is software that advertisers and agencies use to buy digital ad impressions from many sources through one interface, deciding in real time what to bid. Its reported conversions need reconciling with your own analytics.

Formula: Reconciliation ratio = analytics conversions ÷ DSP-reported conversions

What is a demand-side platform (DSP)?

A demand-side platform (DSP) is software that advertisers and agencies use to buy digital ad impressions from many sources through one interface. Instead of negotiating with each website, the buyer sets a budget, an audience and a goal, and the DSP decides in real time which impressions to bid on and how much to pay. In IAB Tech Lab's OpenRTB specification, DSPs are the bidders that act for advertisers, opposite the supply-side platforms that represent publishers.

A DSP is a buying tool. It does not own ad space, and it does not know what happens on your website unless you tell it. It reports the conversions it can see, and those reports are the part that needs checking against your own analytics.

If you saw "DSP" in an audio or engineering context, it can also mean digital signal processing. This page covers the advertising meaning.

Where does a DSP sit among the other ad-tech terms?

TermWorks forWhat it does
Demand-side platform (DSP)Advertisers and agenciesFinds impressions, decides bids, serves the ad
Supply-side platform (SSP)PublishersOffers ad slots to many buyers and collects the winning bid
Ad exchangeNeither sideRuns the auction for each impression between buyers and sellers
Ad networkBoth, as a middlemanBundles many sites' space and sells it as a package
Ad serverEither sideStores the creative, picks the ad and counts delivery

The OpenRTB specification defines the exchange as a service that conducts an auction among bidders per impression; a DSP is one kind of bidder. The mechanics are in real-time bidding and the wider picture in programmatic advertising.

What does a DSP actually do for a campaign?

In practice a DSP bundles five jobs:

  1. Inventory access. One login reaches many exchanges and sellers.
  2. Targeting. Picks impressions by context, location, device, time or audience lists the advertiser supplies.
  3. Bidding. Sets the price per impression within the budget, usually with an automated strategy.
  4. Pacing and frequency control. Spreads spend over the flight and limits how often one device sees the ad.
  5. Reporting. Counts impressions, clicks, completions and the conversions it can attribute.

The fifth job is where the trouble starts, because "conversion" in a DSP report is defined by the DSP.

Why do DSP conversions not match your analytics?

Conversions reported by a DSP and conversions counted in your website analytics differ for predictable reasons:

  • Different windows. A DSP may credit a conversion up to 30 days after a click or a view; a session-based report credits the visit that converted.
  • View-through credit. The DSP counts people who saw but never clicked; your site only sees the visit that happens later.
  • Different identity. The DSP counts devices or cookies; analytics counts people, and the same person on two devices is two for one, one for the other.
  • Overlap across channels. If three platforms each claim the same order, the sum of claims exceeds the real number of orders.
  • Invalid traffic. Bots and accidental clicks count as clicks in one system and are filtered out in the other.
  • Time zones and cut-offs. A conversion at 11:40 pm can land on different days in each report.

A worked example: reconciling DSP reports with analytics

Sample data: three campaigns in one month. The DSP reports conversions; analytics counts conversions in sessions that carry the campaign's UTM tags.

CampaignDSP-reported conversionsAnalytics conversions (tagged)Analytics ÷ DSP
A, prospecting48030062.5%
B, retargeting22019086.4%
C, video954042.1%
Total79553066.7%

Read the table by row, not by total. Retargeting (B) matches closely because it reaches people already on the site, who click and convert in the same session. Video (C) matches least because much of its credit is view-through: people who watched and later searched for you.

Now the cross-platform check. The shop recorded 640 orders in the month, but the platforms (this DSP plus a social platform and a search platform) together claim 1,150 conversions. Claims ÷ actual orders = 1,150 ÷ 640 = 1.80. The same orders are being credited more than once; whichever number you quote, no single report holds the truth.

A reconciliation rule that works: pick one source of truth for counts (your analytics or order system) and use each platform's report only for delivery and direction. Judge a DSP by the ratio of its conversions to your own, tracked over time; a ratio that moves a lot is a tagging or window change.

What should you ask before trusting a DSP report?

Five questions, answered in writing before the flight starts:

  1. What is the attribution window, and does it include views? Write down both numbers (for example "7 days after a click, 1 day after a view").
  2. Is invalid traffic filtered, and by whom? If the DSP filters bots and your analytics filters them differently, expect a gap.
  3. Which time zone are the daily numbers cut in? Match it to your analytics.
  4. How are fees shown? Spend can be reported with or without platform and data fees; compare like with like.
  5. Which conversion event is being counted? A purchase, a sign-up and a page view are very different goals; use the same event on both sides.

Common mistakes with DSP reports

  • Adding platform reports together. The total will exceed reality when channels overlap.
  • Comparing a view-through number to a click-through number. They answer different questions.
  • Changing the attribution window mid-flight. The report will jump without anything else changing.
  • Not tagging the landing page. Without UTM parameters your analytics cannot split DSP visits from others, so there is nothing to reconcile.
  • Treating DSP cost per conversion as final. It uses the DSP's conversion count; recompute with yours.

How to reconcile a DSP in TapCub

Take the three campaigns above. TapCub does not buy, bid on or run ads and is not a DSP; it counts what reaches your site and lets you line that up with the DSP's report.

  1. Tag every DSP campaign with utm_source, utm_medium and utm_campaign (the campaign name from the DSP). See name UTM parameters.
  2. Read tagged conversions. Web analytics › Sources › UTM lists sessions per campaign; click a row to filter and open Web analytics › Goals to count the people who converted.
  3. Check for several credits to one conversion. In All-platform analytics › Behavior › Behavior analysis choose several attribution models side by side (first, last, linear, position-based, time-decay), each as its own column; a channel that is strong in first and linear but weak in last is assisting rather than closing. See attribution and compare attribution models.
  4. Import spend by campaign in All-platform analytics › Business › Channel ROI and compare TapCub's cost per acquisition with the DSP's own.
  5. Keep the comparison dates aligned by checking the site's time zone (see choose the site time zone).

TapCub counts the visits and conversions on your site; impression and view-through data stay in the DSP.

See also: What is a supply-side platform (SSP)? Definition and auction example.

Sources

Was this explanation clear?

We use your answers to fix and expand this page.

Oliver Reed

Product Experience Specialist

Explores usability, user journeys, and feature adoption through practical examples of onboarding and everyday product use.

FAQ

Questions people ask

Short answers to the follow-up questions that come up most.

Still have a question?

Chat with the team behind TapCub — we usually reply within a few hours on business days.

What is a DSP in advertising?

A DSP, or demand-side platform, is software that advertisers and agencies use to buy digital ad impressions from many sources through one interface. It decides in real time which impressions to bid on and how much to pay, then serves the ad and reports results.

What is the difference between a DSP and an SSP?

A DSP works for advertisers and buys impressions; a supply-side platform works for publishers and offers their ad slots to many buyers. They are opposite sides of a programmatic transaction, with an exchange or auction in between.

Why do DSP conversions differ from my analytics?

Windows, view-through credit, identity (devices versus people), channel overlap, invalid traffic and time zones all differ. Pick one source of truth for counts, tag every campaign with UTM parameters, and track the ratio between the two over time.

Can a DSP show conversions your website never saw?

Yes. A DSP can credit people who saw an ad and later converted without clicking, a view-through conversion. Your website sees only the visit that happened. Both can be true, but they answer different questions.

See this number on your own site

Add one script tag for web and apps, and the first reports fill in within minutes.

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics