Influencer marketing
How to measure an influencer campaign: from reach to profitability
Measure an influencer campaign against its objective using agreed definitions and costs. Distinguish interactions, attributed sales and sales caused by the campaign: links and codes support tracking but do not alone demonstrate incremental sales.
A report can contain impressive viewing figures without showing whether a campaign met its objective. Interpretation requires an initial question, a source for each number and a basis for comparison. It also requires acknowledging what the data cannot establish.
Why attributing a sale does not prove causation
Gordon and colleagues (2019) compared measurement approaches across 15 Facebook advertising experiments, comprising 500 million user-experiment observations and 1.6 billion impressions. Observational methods often did not reproduce randomised experimental effects. The study concerns advertising, not influencers. Its relevance here is methodological: exposure and purchasing can be related without all purchases being caused by exposure.
Reference: Gordon et al., 2019
Choose metrics that answer the objective
Specify each definition before comparing results. Reach generally refers to unique accounts under platform definitions, while impressions count exposures. Adding creator reach does not give a deduplicated audience.
For traffic, review clicks and destination sessions. For enquiries, define a valid request. For sales, document returns, discounts, the observation period and attribution rules. State when only platform data is available.
| Objective | Suggested metric | Limitation to record |
|---|---|---|
| Awareness | Reach and impressions | Audience overlap across posts or creators |
| Interaction | Comments, saves or shares | Not equivalent to purchase intention |
| Traffic | Clicks and identified sessions | Platform and destination measurement differences |
| Enquiries | Valid requests and cost per request | Validity criteria and duplicates |
| Sales | Attributed orders and revenue | Window, returns and non-causal attribution |
Define denominators and costs before calculating
Two engagement rates may be incompatible if one uses followers and the other reach. State the formula and apply it consistently. This guide suggests interactions divided by reach where comparable data is available; this is not a universal standard.
Cost per enquiry divides the included cost by valid enquiries. ROAS divides attributed revenue by defined advertising spend; it does not measure profit. Profitability assessment must account for product or service costs, fees, production, amplification and other relevant expenses without double-counting.
Hypothetical example: revenue does not establish profitability
Assume a total campaign cost of €3,000 and €6,000 of attributed revenue after returns. At a 40% contribution margin before campaign costs, those sales would contribute €2,400. Subtracting campaign costs leaves −€600. Under these assumptions, the calculated return relative to campaign cost is −20%.
These are illustrative figures, not ALKATECH results. They also rely on attributed sales: we do not know how many would have happened without the campaign. The calculation should not be presented as incremental ROI. Revenue, margin and causation answer different questions.
Use links and codes with explicit interpretation
Prepare campaign- and creator-tagged links, check the destination and agree an observation period. A code can spread beyond the original audience; someone may see content and purchase later without using it. Explain these limitations.
Measurement also depends on access and consent choices. Missing data should not be interpreted as zero sales, and all sales during the campaign should not automatically be attributed to it.
Finish with a decision and its uncertainty
Compare the objective with available results and costs. Identify what deserves another test, what should change and what remains unknown. Where volume and design allow, consider an experimental comparison with methodological support. A simple before-and-after comparison can be affected by seasonality and other activities.
A useful dashboard keeps definitions, sources, periods and limitations beside each metric. This supports decisions without claiming more precision than the data provides.
Frequently asked questions
What is a good engagement rate?
It depends on platform, format, audience and formula. Compare equivalent data and the campaign objective; this guide establishes no universal threshold.
Does a discount code prove additional sales?
No. It identifies orders using the code but does not alone show how many would have occurred without the campaign.
Can I calculate profitability using only revenue and creator fees?
That would be incomplete if other relevant costs are missing. Define margin, included costs, returns and attribution limitations first.
Sources and scope of evidence
Research figures describe the samples and conditions stated, not ALKATECH results. The examples and templates in this guide are illustrative.
Gordon, B. R., Zettelmeyer, F., Bhargava, N., & Chapsky, D. (2019). A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebook. Marketing Science, 38(2), 193–225. DOI: 10.1287/mksc.2018.1135.
Read the research — Gordon et al., 2019