Own your metrics and centralize every vendor export in one place: that single action lets you reconcile conflicting vendor reports and put a dollar figure on the gap. It matters because 80% of dealers do not own their digital metrics, which leaves vendors grading their own work. Start by sending the checklist below.
TL;DR:
- Request daily campaign exports, spend feeds tied to paid invoices, CRM sales records, and call logs, each carrying IDs and timestamps for matching.
- Before comparing results, define conversions and attribution windows, standardize reporting periods and time zones, and require observed counts to appear separately from modeled figures.
- Use one attribution rule and reporting window across vendors, enforce campaign naming conventions, and assign each sale a canonical CRM field to prevent duplicate claims.
- Normalize identifiers, compare results by campaign and channel, then verify sampled conversions against CRM or point of sale records before calculating each discrepancy’s dollar impact.
Table of Contents
- What should you request from every vendor this week?
- Which data inputs and fields make reconciliation possible?
- Which metric definitions should you lock down first?
- Why do vendor reports conflict, and what fix should you order?
- How do you run the reconciliation?
- How should you present the results to leadership?
- Why independent reconciliation produces better decisions
- How AutoROIQ helps with vendor reconciliation
- FAQ
- Sources
What should you request from every vendor this week?
Send one written request to every marketing vendor and to your internal teams. Vague asks produce vague reports, so name the files, the fields and the owner. Here is what to demand:
- Daily campaign CSV exports with spend, impressions, clicks and conversions, by campaign and ad group
- Monthly spend feeds that tie to the invoices you actually paid
- CRM lead and sales exports, including order IDs and timestamps
- Call and chat logs with call IDs, timestamps and matched lead IDs
- A written description of any modeling, sampling, projection and filtration rules applied to the numbers, which is the kind of disclosure the IAB and MRC attention measurement guidelines call for when vendors report on attention metrics
- Observed counts kept separate from modeled or extrapolated values in every file
Assign ownership before the files arrive. One person centralizes the files, one person maps the fields, and one executive signs off on the final reconciled numbers. Without a named sign-off owner, disputes drift back to whichever vendor argues loudest.
Pro Tip: Give vendors a firm return date and say that reports without raw exports will not count toward the review.
Which data inputs and fields make reconciliation possible?
Reconciliation fails when the datasets cannot be matched. Collect each dataset with the fields that let you join it to the others.
- CRM and sales data: timestamps, sale type, order ID, and a campaign ID or lead ID for matching. MRC guidance on outcomes data treats these datasets as the direct measure of sales and expects vendors to disclose sourcing, recency, cleaning and known limitations.
- Daily campaign exports: spend, impressions, clicks, conversions, campaign and ad group IDs, and the naming convention in use. OEM programs such as the Subaru Certified digital advertising requirements show how far standardization can go, with set ad group names, daily CSV files and query-string tags.
- Verification exports: viewability logs, click logs, and invalid traffic flags. The IAB click measurement guidelines cover what counts as a valid click and how filtration should be reported.
- Web analytics events: a conversion dictionary that says which events count at the campaign level. Industry advice to demand campaign-based conversions across channels applies here.
- Contact center logs: call timestamps, call IDs, matched lead IDs and disposition codes.
If a vendor cannot supply a join key, record that as a finding. It is a measurement gap you now own.
Which metric definitions should you lock down first?
Two reports can both be accurate and still disagree because each uses a different definition. Settle these before you compare a single number.
- Impression buckets: require viewable, non-viewable and undetermined impressions to be reported separately, along with the Measured Rate, as the MRC viewable impression guidelines recommend.
- Conversion definition: list the exact events that count as a conversion and the attribution window that applies to them.
- Observed versus modeled: insist that projections and extrapolations appear in separate columns from observed counts.
- Traffic filtration: require a written account of the audit steps used to remove invalid or fraudulent traffic.
Put the definitions in a one-page glossary and attach it to every vendor request. Vendors then report against your terms, not theirs.
Pro Tip: Ask each vendor to restate last month's report using your glossary. The size of the change shows you how much of the conflict was definitional.
Why do vendor reports conflict, and what fix should you order?
Most discrepancies trace back to a handful of causes. Each has a fix you can require now.
- Attribution model differences: mandate one reconciliation attribution rule for the review period, applied to every vendor.
- Time window and timezone mismatches: standardize reporting windows and document the timezone each system uses.
- Naming and mapping problems: enforce a required campaign naming schema and keep a mapping table, following the discipline shown in the Subaru Certified requirements.
- Sampling, projection and aggregation: require raw counts, with extrapolations segregated and labeled.
- Duplicate conversion counting: define canonical deduplication logic and a source-of-truth field, so a single sale is not claimed by three vendors.
Our own guide on four fixes to stop conflicting marketing reports covers quick wins you can order before the full review begins.
How do you run the reconciliation?
Work in order. Skipping the early steps is what makes late results indefensible.
- Centralize and secure all vendor exports and CRM or POS dumps in one repository with controlled access.
- Normalize fields and map campaign IDs, channels and conversions into one canonical schema. Our guide to fixing dealer website metrics walks through this normalization.
- Reconcile at the campaign and channel level, and flag deltas by dimension.
- Sample-match reported conversions back to CRM or POS records to verify attribution claims, consistent with the CRM dataset guidance in the MRC outcomes standards. Closed loop attribution explains the matching logic.
- Calculate the dollar impact of each discrepancy, write remediation recommendations, and set a cadence for follow-up.
- Institutionalize the changes: naming conventions, a standing reporting rhythm, and an early-review trigger that checks viewability early in the campaign, per the MRC viewability guidance. A working reporting cadence keeps that trigger from lapsing.
Normalized identifiers pay off beyond advertising. Inventory and workspace tools such as Carvia show why consistent vehicle and campaign metadata make records easier to join.
Pro Tip: Run step 4 on a sample of conversions per vendor first. Patterns in the mismatches usually point to the root cause faster than a full audit.
How should you present the results to leadership?
Owners and executive teams want a decision, not a data dump. Build a one-page scorecard with these elements:
- The headline dollar delta and the percent delta between reported and verified results
- A confidence level for the finding
- The root cause, stated in plain terms
- A recommended action with an owner and a date
- Observed counts shown apart from modeled adjustments, in line with IAB and MRC disclosure expectations
- A materiality threshold, so everyone knows which gaps trigger an investigation and which can be noted and monitored
Attach a short appendix with the sample rows used for verification and the mapping table. That appendix is what makes the scorecard defensible when a vendor pushes back.
Why independent reconciliation produces better decisions
The conflict is structural, not an accusation. An independent scorecard settles disputes because no one in the room sells media. The usual outcomes of clearing these conflicts are better budget allocation and stronger vendor contract terms.
— AutoROIQ
How AutoROIQ helps with vendor reconciliation
We do not sell advertising, manage campaigns or take vendor commissions, so our analysis answers only to your numbers. Our Marketing Intelligence Review gives you reconciled datasets, an executive scorecard and remediation recommendations. Our ongoing advisory services keep the process running after the first review. To start, request a Marketing Intelligence Review or contact our advisory team through Autoroiq.
FAQ
How long does vendor report reconciliation take?
The time a focused review takes depends on how many vendors you use and how clean your CRM data is. Most of that time goes to collecting exports and mapping fields, not to the final math.
What data do I need to reconcile vendor reports?
You need vendor campaign exports, spend feeds, CRM lead and sales records, and call or chat logs. Each needs a shared key, such as a campaign ID or lead ID, so records can be matched.
Why do two vendors claim credit for the same sale?
Each vendor applies its own attribution model and time window, so one sale can be counted twice. A single attribution rule and a source-of-truth field in your CRM remove the overlap.
What is the first thing I should ask a vendor for?
Ask for raw daily exports with modeled or projected values labeled separately from observed counts. IAB and MRC guidance expects this kind of disclosure when such metrics are used.
Sources
- NADA Analytics | NADA
- Digital Audience-Based Measurement Standards
- IAB and MRC Attention Measurement Guidelines
- MRC Viewable Ad Impression Measurement Guidelines