{"id":1067,"date":"2026-08-31T18:31:39","date_gmt":"2026-08-31T23:31:39","guid":{"rendered":"https:\/\/www.seomos.com\/?p=1067"},"modified":"2026-09-17T16:08:20","modified_gmt":"2026-09-17T21:08:20","slug":"que-es-un-kpi","status":"publish","type":"post","link":"https:\/\/www.seomos.com\/en\/blog\/que-es-un-kpi\/","title":{"rendered":"What a KPI Is, How to Choose One, and Examples by Goal"},"content":{"rendered":"
\n

A KPI is a key performance indicator<\/strong>A KPI is a measure chosen to evaluate progress toward an important goal. It's not enough for it to simply appear on a dashboard. It must have a definition, a target, a data source, a frequency, and a person responsible for taking action. In marketing, the right KPI connects activity and business results without confusing visits, leads, and customers.<\/p>\n<\/div>\n

What does KPI mean and why does it matter?<\/h2>\n

KPI comes from key performance indicator<\/em>. The Tableau documentation<\/a> It describes it as a measure that allows you to evaluate how a company is progressing toward key objectives. The key is in "key": an organization can record hundreds of metrics and only need a few to guide a specific decision.<\/p>\n

Imagine a company that wants to generate suitable leads for its sales team. It can measure impressions, clicks, visits, forms, conversations, proposals, and contracts. All of these provide information, but not all of them alone indicate whether it's achieving its goal.<\/p>\n

A KPI helps focus the conversation: what we were trying to achieve, what happened, what explains the difference, and what we will do next. If a number doesn't change any decisions, it should perhaps remain as a diagnostic metric and not be the focus of the report.<\/p>\n

Difference between objective, KPI, metric and goal<\/h2>\n
\n\n\n\n\n\n\n\n\n\n
How to turn a general intention into a usable measurement<\/caption>\n
Element<\/th>\nFunction<\/th>\nHypothetical example<\/th>\n<\/tr>\n<\/thead>\n
Aim<\/th>\nExpress the desired result.<\/td>\nIncrease suitable business opportunities.<\/td>\n<\/tr>\n
KPI<\/th>\nMeasure the priority progress towards that result.<\/td>\nQualified applications received per month.<\/td>\n<\/tr>\n
Goal<\/th>\nSet a level and a deadline.<\/td>\nIncrease from 80 to 100 per month by the end of the quarter.<\/td>\n<\/tr>\n
Diagnostic metric<\/th>\nIt helps to understand why the KPI changes.<\/td>\nPercentage of applications that meet the profile.<\/td>\n<\/tr>\n
Control indicator<\/th>\nMonitor to ensure that no other relevant condition deteriorates.<\/td>\nCost per qualified application and response time.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

The classification depends on the context. Response time can be a control indicator for acquisition and the primary KPI for a customer service project. Visits may be relevant for an outreach initiative, but insufficient to demonstrate sales.<\/p>\n

Don't confuse a task with an outcome either. "Publish eight articles" describes an activity. "Get relevant leads from informative content" describes a desired effect. You can track both, but completing the first task doesn't automatically produce the second result.<\/p>\n

\"Relaci\u00f3n
The objective provides direction; the KPI provides evidence; the goal allows for comparison; and the decision transforms the report into management.<\/figcaption><\/figure>\n

How to choose a KPI: a six-step process<\/h2>\n

1. Define the outcome and the population<\/h3>\n

Write down what you want to change and for whom. \u00abImprove marketing\u00bb doesn\u2019t identify a decision. \u00abIncrease qualified requests for the main service in Colombia\u00bb better defines the task. If there are segments with different sales cycles, avoid treating them as if they will respond the same way.<\/p>\n

2. Formulate the question that the indicator should answer<\/h3>\n

For example: Are we generating enough suitable opportunities with the resources available? This question allows us to discard flashy metrics that don't help answer it. It also forces us to define what "suitable" means before reviewing results.<\/p>\n

3. Document the formula and its exclusions<\/h3>\n

If you choose a rate, write the numerator and denominator. "Conversion" could mean forms per session, purchasing users per user, or sales per lead. These are different calculations. Document duplicates, spam, cancellations, and attribution rules to ensure two people get the same number.<\/p>\n

4. Establish a baseline and an explainable goal<\/h3>\n

Use a period that is sufficiently representative of the process. Consider seasonality, budget, and operational capacity. The goal should be linked to a business hypothesis or need; choosing a percentage because it sounds ambitious doesn't help with planning.<\/p>\n

5. Assign responsibility, frequency, and source<\/h3>\n

One person should be responsible for the review and coordinate the actions, even though the outcome depends on several areas. This person defines who maintains the data and who makes the decisions. Daily reviews can help detect errors; evaluating effectiveness may take weeks or months, depending on the cycle.<\/p>\n

6. Define what you will do in the face of a change<\/h3>\n

Write a research rule: if the proportion of qualified applications drops, review queries, ads, and reasons for rejection. Don't turn every variation into an automatic stop-campaign order. First, check data quality, sample size, and context changes.<\/p>\n

\"Proceso
The indicator's data sheet must be able to be reconstructed before designing the dashboard that will display it.<\/figcaption><\/figure>\n

KPI sheet: the fields that should be written down<\/h2>\n

A brief summary avoids recurring discussions about the meaning of the number. This example is a working proposal that you should adapt to your operation; its goals are not universal references.<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Hypothetical form of qualified applications<\/caption>\n
Field<\/th>\nDefinition of the exercise<\/th>\n<\/tr>\n<\/thead>\n
Aim<\/th>\nGrow in opportunities suitable for the main service.<\/td>\n<\/tr>\n
KPI<\/th>\nMonthly number of qualified and deduplicated applications.<\/td>\n<\/tr>\n
Rating criteria<\/th>\nThree criteria are reviewed using the same protocol: compatible need, market served, and valid commercial request.<\/td>\n<\/tr>\n
Baseline and goal<\/th>\n80 per month; reach 100 by the end of the quarter.<\/td>\n<\/tr>\n
Main source<\/th>\nCommercial record reconciled with forms and conversations.<\/td>\n<\/tr>\n
Responsible<\/th>\nMarketing Manager, with review of the sales team.<\/td>\n<\/tr>\n
Frequency<\/th>\nWeekly diagnosis and monthly closing with complete data.<\/td>\n<\/tr>\n
Controls<\/th>\nMaximum cost per qualified request of 150,000 COP and agreed response capacity.<\/td>\n<\/tr>\n
Exclusions<\/th>\nSpam, internal tests and duplicates, while maintaining traceability.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Add the cut-off date, definition version, and change history. If you modify the rating criteria, mark a comparability break. The same label may hide a different series after a form or business procedure is changed.<\/p>\n

Example of a marketing KPI: more forms is not always better<\/h2>\n

Hypothetical exercise, unrelated to SEOMOS results.<\/strong> A company allocates COP 12,000,000 monthly to the acquisition under analysis. It decides to measure qualified applications, with a target of 100 per month and a limit of COP 150,000 per qualified application. We are not calculating Customer Acquisition Cost (CAC): not all of these applications have yet resulted in customers.<\/p>\n

\n\n\n\n\n\n\n\n\n\n
Three fictitious periods with the same cost and different initial business results<\/caption>\n
Indicator<\/th>\nBase period<\/th>\nPeriod 2<\/th>\nPeriod 3<\/th>\n<\/tr>\n<\/thead>\n
Total applications<\/th>\n400<\/td>\n500<\/td>\n420<\/td>\n<\/tr>\n
Qualified applications<\/th>\n80<\/td>\n75<\/td>\n105<\/td>\n<\/tr>\n
Qualified proportion<\/th>\n20 %<\/td>\n15 %<\/td>\n25 %<\/td>\n<\/tr>\n
Cost included<\/th>\n12,000,000 COP<\/td>\n12,000,000 COP<\/td>\n12,000,000 COP<\/td>\n<\/tr>\n
Cost per qualified application<\/th>\n150,000 COP<\/td>\n160,000 COP<\/td>\n114,286 COP approx.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

In period 2, the number of forms increased by 25, but qualified applications fell from 80 to 75. The qualified proportion dropped from 20 to 15: a reduction of five percentage points<\/strong>, equivalent to 25 % of relative drop. These are two different ways of describing the change.<\/p>\n

The team reviews queries, messages, traffic sources, and business criteria. In period 3, 105 qualified requests appear, and the unit cost drops to approximately 114,286 COP. The main KPI exceeds the target, and cost control remains within limits.<\/p>\n

This allows us to investigate what changed, but it doesn't prove that a specific modification caused the entire improvement. Channel mix, seasonality, and other interventions could all play a role. Furthermore, we still need to observe how many opportunities are closing and what contribution they make. The input indicator doesn't replace the final result.<\/p>\n

\"Comparaci\u00f3n
The period with the highest total number of applications is not necessarily the one that generates the most qualified applications. The definition of the KPI changes the interpretation.<\/figcaption><\/figure>\n

Examples of marketing KPIs according to objective<\/h2>\n

Select indicators based on the decision you need to make. This matrix suggests possible relationships; it does not indicate that all indicators must appear together in every company.<\/p>\n

\n\n\n\n\n\n\n\n\n\n
Objectives and families of indicators to guide a dashboard<\/caption>\n
Aim<\/th>\nCandidate KPI<\/th>\nComplementary diagnosis<\/th>\n<\/tr>\n<\/thead>\n
Capture adequate demand<\/th>\nQualified applications from the target market.<\/td>\nOrigin, qualified proportion and reasons for discarding.<\/td>\n<\/tr>\n
Acquire sustainable customers<\/th>\nfull CAC within a defined scope.<\/td>\nClosures, trading duration and contribution.<\/td>\n<\/tr>\n
Improve advertising efficiency<\/th>\nROAS interpreted together with margin.<\/td>\nPurchase value, returns, and attribution.<\/td>\n<\/tr>\n
Facilitate a web action<\/th>\nCompletion rate of the chosen process.<\/td>\nErrors, dropouts, and differences by device.<\/td>\n<\/tr>\n
Increase the value of the relationship<\/th>\nContribution per customer over a fixed horizon.<\/td>\nRepurchase, tenure and service cost.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

For formulas and their limits, consult the guides of CAC<\/a>, ROAS<\/a> y customer lifetime value<\/a>. Keeping them separate allows this board to focus on decisions and not on an endless collection of acronyms.<\/p>\n

In SEO, Search Console provides clicks, impressions, CTR, and average position based on your performance report documentation<\/a>. These are search signals, not sales figures. To assess commercial demand, you must relate them to what's happening online and in the sales process, without claiming a level of attribution you don't have.<\/p>\n

On social media, the usefulness of reach and engagement depends on the objective. You can track discovery and relevant inquiries within a strategy of SEO for social media<\/a>, keeping exhibition, conversation and confirmed sale separate.<\/p>\n

How to connect KPIs with reliable data<\/h2>\n

Define a primary source for each fact. The ordering system confirms purchases; the business record defines opportunity statuses; the advertising platform reports investment; web analytics monitors events. No single tool should be considered the sole source of all answers by default.<\/p>\n

Google Analytics allows you to mark as key events<\/a> actions that are especially relevant to the business. Marking an event doesn't automatically make it a complete KPI: you still need to decide what objective it measures, how it's measured, and what action it triggers.<\/p>\n

To understand where an action is lost, the GA4 funnel explorations<\/a> They allow you to observe defined steps. Verify their sequence, entry conditions, and the unit counted. A user funnel should not be compared without adjustments to a count of repeated events.<\/p>\n

If the objective depends on a landing page<\/a>, Test your measurement from the form to the business record. Exclude internal tests from operational reports and ensure that a retry does not generate two attempts. Missing data must be identified as such, not simply rendered as zeros.<\/p>\n

How to review a dashboard and turn it into actions<\/h2>\n

Organize the review around four questions: Did we meet the goal? Is the data comparable? Which segments explain the variation? And what action will we test? Accompany each KPI with current value, target, period, trend, and update date.<\/p>\n

Color helps to pinpoint a signal, but it doesn't replace explanation. A value in red could be provisional data, a measurement problem, or a real deviation. Before changing your budget, distinguish between these possibilities. Also, don't average rates from groups of different sizes: consolidate their numerators and denominators.<\/p>\n

Record the decision, a responsible person, and a date to evaluate the outcome. In a strategy of SEO positioning<\/a>, The timeframe should allow for observation of demand and changes in visibility. paid media<\/a>, Also review conversion delays and operational limits. The frequency of monitoring data doesn't have to be the same as the frequency of intervention.<\/p>\n

Mistakes to avoid before presenting results<\/h2>\n