A redesign makes comparison difficult if nobody records how the old site works. After launch, a changed form, traffic mix or routing rule can be mistaken for a design result because the original evidence was never preserved.
TLDR: Document the existing lead-capture path, report definitions and known defects before design begins. Preserve observed data with its date range and limitations. Use the baseline to test continuity and interpret later changes, without treating a before-and-after difference as proof of uplift.
We recommend mapping each important entry page to its CTA, form or booking action, confirmation step and CRM owner. Include alternate routes such as a header button or a form reused on several pages.
Record the actual destination and asset identifier where available. “Contact form” is not specific enough when several forms share a similar name. Note any external tool or integration that participates in the path.
The landing-page guide can help explain the page's purpose. The baseline should show what the current implementation does, including the parts that do not work well.
As checked on September 7, 2026, HubSpot's single-form reporting documentation defines conversion rate as submissions divided by form views. It notes that API submissions can count as submissions without generating form views. That distinction matters when comparing forms or interpreting a rate. HubSpot form performance definitions
We recommend recording the report, filters, date range, time zone and unit of measurement with the export. Separate submissions from unique people and from sales-qualified outcomes. Do not assume one form submission equals one new customer opportunity.
Keep unavailable data blank and labeled unknown. A missing tracking period should not become a zero-conversion period in the baseline.
This example is synthetic. It describes the fields to capture, not an observed website result.
| Baseline item | Example record | Limitation to retain |
|---|---|---|
| Primary path | Service page to consultation form to confirmation page | Record alternate entry paths separately. |
| Asset identity | Form F-EXAMPLE-03 reused on two pages | Form-level totals may span both pages. |
| Owner route | Requests go to the service-review role | Confirm actual assignment evidence. |
| Measurement period | An agreed pre-redesign observation window | Dates and traffic exclusions are not yet supplied. |
| Known issue | A synthetic test found an unclear confirmation message | This is an example defect, not a real site finding. |
If you use a worked rate, show the denominator. For example, 20 synthetic submissions divided by 500 synthetic form views is 4%. That arithmetic is not a benchmark, a target or evidence that the redesign will improve performance.
We recommend retaining an approved test of each important path. Capture what the visitor sees after submitting, which information is stored, who receives the request and whether the expected internal next step appears.
Use synthetic test details and an authorized process. A baseline test can otherwise trigger production notifications, enrollments or marketing actions. Document the test's limits before executing it.
Review the specific submission and associated record where appropriate. HubSpot supports inspecting individual form submissions and their response details; the necessary permissions still apply. HubSpot submission review
Do not preserve a bad behavior merely because it exists today. We recommend classifying each item as intended behavior to retain, known defect to fix or unresolved question to investigate.
This distinction lets design decisions improve the experience while protecting the critical operating path. It also prevents a launch team from calling an inherited problem a new regression without checking the record.
A baseline should let the new-site team understand the current lead management process without guessing. Record the entry point, required information, CRM result and receiving owner.
Include the sales and marketing teams that use the captured information. The marketing team may need to know which offer attracted the inquiry, while the sales team needs a reliable next action and enough context to respond.
Do not define success solely as more leads. Agree on the qualification rule and the operational path before choosing a performance comparison.
For each HubSpot form or external form, record the fields, required status, labels and business reason. Include hidden values or conditional behavior where present in the approved implementation.
A company-size field might support qualification. A service-region field might control the receiving team. Removing either could change the meaning of the later lead report.
This does not mean every existing field should remain. Classify each as necessary now, suitable for later collection or requiring a decision. Preserve the reason for an approved change.
A redesigned form can look the same while using a different asset or downstream configuration. Record the form identifier and the pages that reuse it.
Include confirmation messages, booking routes, notifications and any connected workflow that affects the inquiry. A screenshot alone will not preserve those relationships.
When an asset is replaced, document the old and new identifiers. That relationship helps explain changes in reporting and troubleshoot a missing follow-up.
Define what makes an inquiry accepted, held or rejected. Keep the criteria understandable to the receiving team.
A form submission is an event. A qualified lead is a business classification applied under an agreed rule. A new contact is a CRM record state. Do not combine them into one baseline total.
Record how the inquiry reaches an owner, including the exception route when required information is missing. A successful confirmation page does not prove the sales process began.
The old and new sites may receive different traffic. Keep the relevant campaigns, source definitions and observation dates with the baseline.
If a new offer, a changed form or a revised qualification rule launches alongside the design, record it. Those changes can affect conversion rates independently of the visual layout.
An unavailable period, missing event or unknown source is not a zero result. Preserve the gap and its effect on the comparison.
Do not reconstruct historical website behavior from the current configuration alone. A form may have changed during the reporting period.
Use the strongest evidence available, state what it covers and retain unresolved questions for the launch review.
Include header CTAs, mobile navigation, reused forms and returning-visitor paths where they matter to the site. Do not assume the primary service-page journey represents the entire site.
Use an approved synthetic test plan and record its effects. Real submissions can trigger notifications or automation, so baseline collection needs the same care as launch testing.
First verify that the new site preserves or intentionally improves the approved lead-capture path. Then evaluate performance over a stated period.
A working route is an acceptance result. A higher conversion rate is a measurement result with its own comparison limits. Keeping them separate makes the redesign review easier to interpret.
We recommend repeating the continuity tests first, then comparing performance over a documented period. Record material changes in traffic, campaigns, forms, qualification rules and tracking alongside the design changes.
A before-and-after difference can have several explanations. Do not attribute the whole change to the redesign without appropriate evidence. A controlled page experiment is a separate decision with its own plan.
If the platform itself is changing, the WordPress and HubSpot CMS comparison addresses the broader choice. For help establishing the baseline and acceptance plan, explore website design services or request a scoped review.