How to Measure Small-Business Website ROI

Start with the business decision you want the website to support. A service business may want qualified estimate requests or booked consultations. A shop may want completed orders. A nonprofit may want a completed registration or donation. Choose an action that represents progress for your organization, not a number that is merely easy to count.

Set the measurement period and write down which costs belong in it. Depending on the question, costs might include design or development, hosting, domain renewal, maintenance, content work, and paid promotion. If the website also supports offline sales, note which part of the journey you can observe and where your records will be incomplete.

Map the actions that can turn a visit into a lead. A contact form should have a clear success state after submission. An appointment tool should show that a booking was completed, not just that someone opened the calendar. A phone call may need a trackable number or a simple question in your intake process. Test each path yourself and confirm that the record reaches the place your team actually checks.

Count completed actions once. A button click is useful for diagnosing a page, but it is not a lead if the form never arrives or the appointment is not booked. Keep a separate count for inquiries that meet your definition of a qualified lead, such as being in your service area and asking about a service you provide.

Attribution is a record of the touchpoints you can observe; it is not proof that one page or campaign caused a sale. Use campaign tags consistently on links you control, keep the original source when your analytics or customer system provides it, and consider adding an optional “How did you hear about us?” question to an inquiry form. A person’s answer is useful context, but it is self-reported and may not capture every interaction.

Pick a rule that matches your tools. First-touch attribution credits the earliest recorded source. Last-touch attribution credits the final recorded source before the inquiry. If you have enough reliable data, you can preserve several interactions instead of forcing one source to represent the entire journey. State the rule next to every report and do not compare two reports that use different rules as if they were the same measure.

A lead count is an early signal, not revenue. Carry the inquiry source into the sales record and update its status as the conversation progresses: qualified, scheduled, proposal sent, won, lost, or still open. Use the same lead identifier where possible so a form submission and a later sale do not become two separate wins.

When a sale closes, record the amount collected and the direct costs needed to deliver it. Contribution profit or gross profit is more useful for ROI than top-line revenue because it accounts for the cost of fulfilling the work. If customers usually return, decide in advance whether the calculation includes only the first sale or a defined period of repeat business; avoid adding future value that has not been observed.

A straightforward calculation is: (attributed contribution profit − website investment) ÷ website investment × 100. The result is a percentage for the period and attribution rule you selected. If the investment is zero, the calculation is undefined; if you do not yet have closed sales, report qualified leads and open opportunities separately rather than presenting an estimated ROI as a measured result.

For example, suppose a business records three closed customers with $700 in contribution profit each during a chosen period. That is $2,100 in contribution profit. If the website costs included in the same calculation total $1,050, the arithmetic is ($2,100 − $1,050) ÷ $1,050 × 100 = 100%. This is a hypothetical example to show the formula, not a benchmark or a prediction for another business.

Be consistent about one-time build costs. You can count the full build cost in the period it was paid, or allocate it over a stated useful period for an internal comparison. Either way, show the rule and include recurring costs such as hosting or maintenance that belong to the same analysis.

Read the path in order: relevant visits, completed actions, qualified leads, booked or active opportunities, and closed business. A gap can help you choose what to inspect next. Few completed actions may point to a broken form or unclear next step. Many inquiries that are not a fit may mean the page needs clearer service or eligibility details. Leads that do not close may need a sales-process review rather than another website redesign.

Compare your own periods using the same definitions, costs, and attribution rule. Do not treat an industry average or a search-volume estimate as a target when it has not been validated for your business. If you change the page, tracking, price, or campaign, make a note of when it changed so you can interpret the next period honestly.

A first pass can be small. In week one, define the primary conversion, the qualified-lead rule, the cost boundary, and the attribution rule. In week two, test every important path and fix missing records. During the remaining weeks, record inquiries and outcomes in one consistent place. At the end of the month, review what is known, what is still open, and what tracking needs attention.

This is a proposed working cadence, not a finding from the national search-intent brief or a promise that one month will produce enough data. A low-volume business may need a longer period. Keep collecting results until the sample is useful for the decision you need to make, and show uncertainty where the record is incomplete.