Objectives and KPIs: from business goals to what communication must change

An objective says what must change. A KPI (key performance indicator) is the number that tells you whether it is changing. Integrated marketing communications needs objectives in three connected layers, and a short chain of KPIs for each.

• Business objectives are about revenue, profit, market share, retention or reputation.
• Marketing objectives are about customers and markets: who buys, how often, at what price, and who stays.
• Communication objectives are about what people notice, know, feel and do because of what the brand said and showed, wherever they met it.
• KPIs follow from the objective, never from whatever a platform reports by default.
• Leading indicators move early and help you steer; lagging indicators move late and prove the result.

The tools for this are old and well tested: the hierarchy of effects published in 1961¹, Russell Colley's DAGMAR from the same year²³, and the evaluation framework that AMEC, the professional body for communication measurement, launched in 2016⁵.

Most measurement arguments start with a missing layer. A team is asked to "grow sales" (a business objective) and comes back with "two million impressions" (an activity count). Nothing in between explains how one is supposed to lead to the other, so nobody can tell whether the campaign worked or why.

Write the three layers as a chain, each one answering "how?" for the layer above:

1. Business objective. What the organisation must achieve: grow revenue in a new market, keep customer churn below a ceiling, recover trust after a product recall.
2. Marketing objective. What must change among customers for that to happen: recruit buyers who have never bought the brand, persuade existing customers to buy more often, defend the price, win back lapsed members.
3. Communication objective. What communication must achieve, with whom: make the brand known among people who buy the category, make one idea understood, shift an attitude, prompt a first trial or a renewal.

Each objective links to a hub with the disciplines that usually serve it, for example brand awareness, customer acquisition, retention and loyalty and reputation and crisis.

Each layer should be written so that someone else can check it. The usual test is SMART: specific, measurable, achievable, relevant and time-bound. Smart Insights, for example, pairs SMART objectives with KPIs for each stage of its RACE planning framework⁹. AMEC's Barcelona Principles, the measurement industry's shared statement of good practice, treat setting goals as the prerequisite for communication planning, measurement and evaluation⁷. The 2025 revision, Barcelona Principles 4.0, adds that objectives should be responsive: they can shape a campaign while it runs, not only judge it at the end⁸.

Two publications from 1961 still shape how communication objectives are written.

The hierarchy of effects. Robert Lavidge and Gary Steiner asked, in the Journal of Marketing, what advertising is supposed to do. They argued that immediate sales, even when they can be measured, are an incomplete test of whether advertising works, and proposed that people move towards a purchase through a series of steps: awareness, knowledge, liking, preference, conviction and purchase¹. The point was practical: if the later steps take time, you need measures for the earlier ones.

DAGMAR. In the same year Russell Colley wrote Defining Advertising Goals for Measured Advertising Results for the Association of National Advertisers (ANA). Its contribution was discipline: an advertising goal should be a communication task, defined for a specific audience and stated so that the result can be measured against it³. Colley's stages are usually summarised as awareness, comprehension, conviction and action³. The approach travelled well beyond consumer brands; a 1987 study in the International Journal of Advertising examined how 88 industrial companies used DAGMAR and the hierarchy-of-effects model to set their objectives².

The warning. The step models are a useful checklist but a poor description of how people actually respond. Robert Heath and Paul Feldwick argued in 2008 that the information-processing view behind these models had been the wrong model of advertising for fifty years, because much advertising works through emotion and low-attention processing rather than step-by-step persuasion⁴. In practice this means two things. First, do not insist that people pass through the steps in order; a person can buy before they can tell you anything about the brand. Second, measure several layers at once (awareness, associations, consideration, behaviour) rather than waiting for one to "complete" before looking at the next.

A good rule is one objective, one short chain: a lagging KPI that proves the objective was met and one or two leading KPIs that tell you early whether you are on track.

Why both? Binet and Field's analyses of IPA effectiveness cases found that brand-building effects grow slowly and last, while sales activation produces short-term effects that fade¹⁴. A plan judged only on lagging sales numbers will cut the slow work; a plan judged only on leading signals may never prove anything. The IPA's own ten-year review of that research argues that short-termism kept growing after it was published, and that effectiveness fell as the share of budgets going to activation rose¹⁵. Leading brand measures such as mental availability, how readily a brand comes to mind in buying situations, can be measured too; researchers at the Ehrenberg-Bass Institute have published an approach for measuring advertising's effect on it¹³.

Each discipline then contributes the metrics it can honestly report, and each needs a definition:

• Digital advertising. An impression is not the same as a chance to be seen. Under the Media Rating Council's guidelines, a display ad counts as viewable when at least 50% of its pixels are in view for at least one continuous second, and a video ad when 50% of its pixels are in view for two continuous seconds¹⁰. In November 2025 the IAB and the MRC published guidelines for measuring attention as well¹¹. See digital advertising.
• Email. Apple's Mail Privacy Protection downloads remote content in the background whether or not the reader engages, which prevents senders from knowing whether a message was opened¹². Use clicks, replies and conversions instead of opens. See email marketing.
• Public relations. AMEC's principles state that advertising value equivalents (AVEs) are not the value of communication⁷. Report coverage among the audiences that matter, the messages that came through, and what changed as a result. See public relations.
• Social, search, events and the rest. Choose the output that sits closest to the outcome you want (saves and shares rather than likes, qualified visits rather than all visits, meetings booked rather than badge scans), and write down how each is measured before the campaign starts.

AMEC launched its Integrated Evaluation Framework in 2016 as a free, interactive tool that puts the Barcelona Principles into practice, step by step⁵. It takes a planner from organisational objectives to communication objectives, then to audiences, plans, benchmarks and targets, and on to activities, outputs, out-takes, outcomes and organisational impact⁶. In 2025 AMEC reordered the Barcelona Principles themselves to follow the steps of the framework⁸.

The framework was written for PR, but its logic is the logic of the whole integrated plan. Outputs differ by discipline: an ad delivers impressions, a press office delivers coverage, a creator delivers posts. Out-takes and outcomes are shared, because they happen in the same audience's minds. That is where an integrated plan joins its paid, owned, earned and shared parts (the PESO model): one set of communication objectives, measured once, fed by every discipline.

1. Write the business objective in one sentence, with a number and a date.
2. Write the marketing objective beneath it: which customers, which behaviour, by when.
3. Write one to three communication objectives: audience, the change you want (know, feel, do) and the deadline.
4. For each, choose one lagging KPI and at most two leading KPIs. Record the definition, the data source and the baseline before launch.
5. Map every discipline's output metrics to the out-take or outcome they are meant to move. Drop any metric that maps to nothing.
6. Retire weak proxies: no AVEs, no email opens as evidence, no impressions without a viewability standard.
7. Set a rhythm: leading indicators reviewed monthly, lagging ones quarterly, with the objective itself reviewed if the market changes.
8. Carry the same objectives into the budget and the measurement plan, so all three documents use the same words.

• Lavidge and Steiner's 1961 article, reprinted online¹, and Heath and Feldwick's critique of the step models⁴.
• AMEC's Integrated Evaluation Framework⁵⁶ and the Barcelona Principles 3.0 and 4.0⁷⁸.
• Binet and Field's ten principles from The Long and the Short of It¹⁴.
• On IMC Institute: measuring effectiveness, the brand platform and message house, and the glossary entries for KPI and brand lift.