A busy marketing team is not evidence of effective marketing
The monthly report arrives with more traffic, impressions and content published. The business owner asks a simpler question: did any of this make growth more likely?
Most reports struggle to answer because they mix three different things. Activity describes what the team did. response describes what audiences did. commercial evidence shows whether the right buyers progressed toward profitable revenue. All three are useful, but they are not interchangeable.
Marketing can create demand before revenue appears, particularly when several people influence the decision or the sales cycle is long. That does not remove accountability. It changes the evidence and time window required. A programme aimed at creating preference should not be judged by this week's direct sales, but it should eventually improve qualified demand, account engagement, sales conversations or the economics of winning customers.
My verdict: marketing is working only when the business can connect an intended customer, a meaningful change in behaviour and a credible commercial consequence. If the only evidence lives inside an advertising or analytics platform, confidence should remain provisional.
The Marketing Proof Stack
I use five layers. Each answers a different owner-level question. Evidence becomes stronger as it moves closer to customer value, but leadership should inspect every layer because a downstream result can hide an upstream weakness.
Business target
What must change: revenue, qualified pipeline, customer volume, market entry or retention?
Intended buyer
Are the right companies or customers seeing and understanding the offer?
Meaningful response
Are they taking actions that signal real interest, not merely producing cheap clicks?
Commercial progression
Do enquiries become accepted opportunities, purchases or credible buying steps?
Economic return
Do won customers create enough contribution, value and cash flow for the investment?
Attribution is a useful model, not a complete history
Attribution means assigning credit to ads, clicks and other factors along the path to a meaningful action. Google Analytics describes it in those terms. The word assigning matters: the report is applying a model to observable events, not reconstructing every conversation, recommendation or decision that influenced the buyer.
Use attribution to compare evidence, find patterns and improve decisions. Do not let a precise-looking percentage override what sales hears, what customers say or what finance records.
Use the Marketing Evidence Reconciliation before accepting the report
The Proof Stack shows what good evidence looks like. The reconciliation tests whether the records agree. I use four ledgers because most false confidence appears where one system hands the customer to another.
What was funded?
Media, people, technology, discounts and production costs by programme and period.
Who responded?
Source, intended buyer, message, first meaningful action and whether the demand was new or already present.
What happened next?
Accepted enquiry, sales contact, opportunity, purchase, rejection reason and time between stages.
What did the business retain?
Revenue, contribution, repeat value, cancellations, returns and the cash timing needed to support growth.
Start with totals, then investigate differences. If advertising reports 100 leads but the CRM holds 73, do not average the numbers or choose the more flattering source. Find whether duplicates, consent rules, failed handoffs, date windows or stage definitions explain the gap. A missing record is a measurement problem before it becomes a performance conclusion.
| Evidence layer | Illustrative record | Owner interpretation | Next decision |
|---|---|---|---|
| Advertising platform | 100 reported leads | Response volume exists; quality is unknown | Do not scale from this number alone |
| CRM receipt | 73 unique records | Twenty-seven reported actions need reconciliation | Repair duplicate or missing-record causes |
| Sales acceptance | 24 suitable conversations | The cost of useful demand is materially higher than cost per form | Feed rejection reasons into targeting and qualification |
| Commercial outcome | 5 opportunities; revenue still maturing | The programme may be creating pipeline, but ROI is not yet proven | Review after the normal sales cycle |
This example demonstrates the method, not a benchmark. Your counts, buying cycle and economic threshold will differ. For lead-generation businesses, Google's qualified-lead and converted-lead categories allow offline CRM stages to describe progress beyond the initial form. For ecommerce, reconcile analytics purchases with the store, refunds and contribution rather than treating attributed revenue as profit.
If the handoff itself is unreliable, review conversion tracking and measurement. If the evidence is sound but weekly decisions are inconsistent, review the paid growth service model and what paid ads management should include.
Use three decision windows instead of one monthly verdict
A monthly report often compares costs incurred now with revenue created by older marketing. That timing mismatch can make a new programme look unproductive and a stopped programme look successful.
Operational evidence
Is tracking reliable? Is the intended audience being reached? Is the proposition generating a meaningful response? Inspect early enough to repair obvious waste.
Commercial evidence
Are qualified enquiries, target accounts or buying actions appearing and progressing? Review over the normal evaluation period, not a convenient calendar month.
Financial evidence
Are customers won at acceptable acquisition cost, payback and contribution? Read cohorts through a complete buying cycle before declaring return.
Define the expected evidence and review date before a programme starts. Otherwise “give it more time” can protect weak work indefinitely, while a premature ROI demand can force every budget into channels that capture existing demand and starve the work that creates future preference.
Google Analytics' key-event attribution paths report shows touchpoints, revenue and the time to a key event; Google Ads attribution reports expose conversion paths and assisted conversions. Those views help explain timing and sequence, but they still represent observed digital evidence rather than every factor in a customer decision. The owner-level lesson is simple: the marketing, sales and finance views need to describe the same cohort.
A seven-question owner scorecard
You do not need another dashboard. Ask these questions in the monthly growth review and require the underlying evidence.
| Question | Evidence to inspect | Decision it supports |
|---|---|---|
| What business outcome was this programme hired to change? | Target, baseline, margin and time horizon | Whether the investment has a defined job |
| Which buyer or customer problem is responding? | Audience, query, account, message and customer language | Whether demand is relevant |
| What happened after the first response? | Sales acceptance, purchase, opportunity and stage movement | Whether attention becomes commercial progress |
| How complete is the evidence? | Missing sources, duplicate leads, offline outcomes and reporting gaps | How much confidence to place in the conclusion |
| Does the review window match the buying cycle? | Cohort dates, deal age, time to purchase and seasonality | Whether the verdict is early, fair or late |
| Could another factor explain the change? | Pricing, sales capacity, stock, product, market and previous campaigns | Whether marketing deserves the credit or blame |
| What will we do differently because of this evidence? | One funded action, owner and review date | Whether the report is decision-useful |
For owners, revenue alone is still incomplete: the economic test should use the value the business retains after relevant delivery costs, discounts, returns or servicing requirements. The right calculation depends on the business model, but it must be defined before a report labels the investment efficient.
Scale, repair, continue or stop from the evidence
Scale carefully
Increase investment in steps and check whether buyer quality, conversion and economics hold.
Repair the system
Investigate offer, qualification, sales follow-up, landing experience or customer friction before buying more reach.
Continue with a deadline
Keep the programme only with a named evidence milestone and a review date tied to the buying cycle.
Stop or redesign
Do not defend activity. Change the buyer, problem, proposition or channel role before spending again.
If traffic is the visible symptom, use the Traffic-to-Revenue Diagnostic. If lead volume is masking weak sales, use the CPL-to-Revenue Truth Chain. If the economics are deteriorating, use the CAC Pressure Map. The Paid Ads Pipeline Calculator can make revenue assumptions explicit.
My paid growth services connect acquisition, conversion, measurement and CRM feedback. You can review documented case studies, see how evidence and limitations are handled, learn how I work directly, or request a marketing evidence diagnostic.
Practitioner note: the first question I ask when a report looks impressive is: “What decision can the owner make from this?” If the answer is only “marketing should keep doing more marketing”, the scorecard is not finished.
A 30-day plan to turn reporting into a growth decision
Define the commercial job
Name the customer, outcome, contribution logic, expected buying cycle and the evidence required at each stage.
Reconcile the records
Match platform actions with analytics, CRM or store records. Document gaps, definitions and dates instead of hiding them in a blended total.
Find the first break
Identify whether the earliest material weakness is buyer fit, response, handoff, sales progression, customer value or evidence quality.
Fund one decision
Choose scale, repair, continue or stop. Assign an owner, expected commercial effect, review date and reversal condition.
Do not rebuild every dashboard at once. The goal is one credible decision and a repeatable review rhythm. If the first break is unclear, the B2B paid media audit framework provides a deeper account-to-CRM diagnosis. If traffic is growing without revenue, use the Traffic-to-Revenue Reconciliation.
Sources and evidence notes
Sources and current search results were checked on 20 September 2026. The Marketing Proof Stack, Marketing Evidence Reconciliation, owner scorecard, decision windows and 30-day plan are original ThomPerformance analysis. The sample reconciliation is explicitly illustrative, not client proof. No search volume, universal marketing ROI benchmark or fabricated performance is used.
Frequently asked questions
What is the best way to tell if marketing is working?
Start with the business outcome, then look for connected evidence: the right audience is responding, qualified enquiries are progressing, customers are being won and the economics fit your margin and cash-flow requirements. No single platform metric can prove all four. The evidence must also be reviewed over a window that matches the buying cycle.
Which marketing metrics should a business owner review?
Review qualified demand, sales acceptance, new opportunities or buying actions, revenue won, customer acquisition cost and payback. Use traffic, clicks, reach and cost per lead as diagnostic measures. The exact scorecard depends on whether the business sells online, through a sales team or through a longer consideration process.
How long should I wait before deciding that marketing is not working?
Wait long enough for the expected customer action to occur, but do not wait to inspect early evidence. Check delivery and response quickly, qualified demand after enough meaningful interactions, and revenue over a complete buying cycle. Define these decision windows before launch so patience does not become an excuse for weak work.
Can marketing be working even if revenue has not increased yet?
Yes, particularly when the sales cycle is long, but the claim needs evidence. You should see stronger demand from the intended buyer, more qualified opportunities, improved progression or credible account engagement. If only impressions and clicks are rising, the business does not yet have enough evidence to call the programme successful.
Should I judge every marketing channel by direct revenue?
No. Some channels capture existing demand, while others create preference or reduce perceived risk. Give each channel a defined job, then judge whether it performs that job and contributes to the wider commercial result. Still require a clear connection to business outcomes; different measurement does not mean no accountability.
Marketing should produce evidence before it asks for more budget
Start with the business outcome. Reconcile the investment, demand, commercial progression and customer value across the right decision window. When those records agree, the business can scale with greater confidence. When they disagree, the earliest material gap tells you what to repair before funding more activity.
Which layer is weakest in your current reporting: buyer fit, evidence quality, commercial progression or economics?