MarketingKPI / metricModelAccessible

Conversion rate

How should conversion rate be measured and interpreted?

AccessibleTacticalOrganisation3 min read

Contents

Helps managers answer: To what extent are we able to convert potential customers into actual customers?

Conversion rate measures the proportion of an eligible audience that completes a defined action. The action may be a purchase, but it can also be a quote request, registration, download or another meaningful step. After a business attracts people to a store, website or sales process, this indicator shows how effectively that attention becomes progress or revenue.

When to use it

Answer the key performance question: “To what extent are we able to convert potential customers into actual customers?”
Assess this KPI within the Marketing and sales perspective.
Plan data collection, formula use, reporting frequency, and data-source requirements for this KPI.
Compare results against the targets, benchmarks, examples, or trend guidance available for this KPI.

Origins

Conversion measurement predates the web in direct-response advertising, mail order and sales-funnel management, where marketers divided desired responses by the audience exposed or contacted. Digital commerce made the measure immediate and granular during the 1990s. Its apparent simplicity hides a design choice: both the qualifying population and the conversion event must be defined, and attribution across multiple contacts must be handled consistently.

What it is

Perspective: Marketing and sales perspective.

Key performance question: To what extent are we able to convert potential customers into actual customers?

A conversion must be defined in relation to the objective. In a physical setting it might be a store visitor who buys an item or asks for a quotation. Online, it could be an order, telephone call, membership registration, newsletter subscription, software download or another response to marketing or content.

The measure exposes whether marketing, selling and operations work together. If 500 people arrive at a site or shop and none buys, the problem may be mismatched expectations, poor navigation, unavailable products, weak service or an unacceptable price. Conversion identifies the break; diagnosis still requires evidence about its cause.

The term covers several related measures, including:

Visitor sales conversion rates
Lead generation conversion rates
Click-through rates
Tender or quote conversion rates

How to use it

Measurement

Data collection method

Match collection to the environment and define the unit before calculating. Web analytics can follow eligible users through the funnel from an initial prompt to a target action such as a purchase. Physical retailers may compare entrance-counter data with transaction counts. More advanced camera and software systems can automate visitor-flow and conversion reporting, subject to privacy and data-protection requirements.

Formula

At its simplest, divide completed goals by eligible visitors and express the result as a percentage:

Calculate the same ratio at successive funnel stages: ad or page view to visit, commonly called click-through rate (CTR); click to populated basket; and basket to completed order. Stage rates reveal where the largest loss occurs.

Frequency

Monitor conversion continuously where the data are available, and compare like-for-like cohorts over time.

Source of the data

Digital measures come from consented event tracking or web-analytics systems. Physical-location measures combine visitor-counting or tracking data with sales records.

Cost/effort in collecting the data

Collection effort varies by setting. Once events and goals are correctly configured, online measurement can be inexpensive and largely automated.

Stores, shopping centres, airports and other physical spaces may require dedicated counters, cameras and reporting software, making implementation materially more expensive.

Target setting/benchmarks

Set targets for the particular product, channel, traffic source, device, market and stage; a generic benchmark can mislead. Figures of 2%–3% are often quoted for online retail, while ClickZ reported rates around 10% or more for retailers such as Amazon and eBay. Physical-store rates are normally expected to be higher because entering a store signals stronger intent. Treat these historical examples as context, not current universal standards.

Example

Consider an online retailer using Facebook banner advertisements. The ad appears to an estimated audience of 1 million, providing the starting population for the funnel.

A 2% click-through rate sends 20,000 people from the banner to the retailer’s site:

Of those 20,000 visitors, 3,000 add products to a basket, producing the following stage conversion:

From the 3,000 populated baskets, 2,300 become purchases. The basket-to-sale rate is therefore:

The retailer also calculates the overall visitor-to-purchase conversion as follows:

In a physical example, a large retailer wanted comparable conversion figures across its store network. Entrance cameras counted people entering and leaving each location, while transaction records supplied the number of completed sales:

Top practical tip

Define the numerator, denominator, attribution window and eligibility rules before comparing results. Segment by funnel stage and traffic source so a change in audience mix is not mistaken for a change in performance.

Top pitfall

Do not divide conversions by unfiltered raw traffic. Bots, spiders, repeated visits and internal activity can inflate the denominator; historical estimates placed non-human traffic between 10% and 30%, with the range often restated as 10%–30%. Use an appropriate unique-user or eligible-session measure and document exclusions.