Key facts
- Lead source attribution, recording which marketing channel actually produced a lead inside your CRM, is what lets you eventually connect ad spend to closed revenue, and it has to be captured at the moment a lead comes in, not added after the fact.
- CRM platforms like HubSpot, Salesforce, and GoHighLevel each have their own data export limitations and API access rules, and a vendor should be asked specifically how portable your data is if you decide to switch CRM platforms later.
- A custom-built CRM can be cheaper to run monthly than a licensed platform at scale, but it depends entirely on ongoing developer support for maintenance and changes, which is a different kind of ongoing cost than a subscription fee.
- Lead scoring, ranking incoming leads by how likely they are to convert, requires enough historical data and clearly defined criteria to be useful, and a vendor promising accurate lead scoring on day one with no historical data to learn from should be asked how that is possible.
- Duplicate lead records, the same person entering the CRM twice through two different forms or channels, are a common data quality problem that a good integration should catch and merge, and a vendor should describe their deduplication approach specifically.
Scope boundaries
State which CRM platform you are using or want to use, HubSpot, Salesforce, GoHighLevel, a custom-built system, or another, since the integration approach and cost differ significantly by platform.
List every source that should feed leads into the CRM: your website's contact forms, phone calls if you use call tracking, chat widget submissions, and any ad platform lead forms (Meta lead ads, for example) that bypass your website entirely.
State whether lead routing and automated follow-up (an email sequence triggered when a lead comes in, a notification to a sales rep) are in scope alongside the raw data connection, since these are a meaningfully larger project than simply piping form data into a CRM database.
What the vendor must be given
Give the vendor admin access to your CRM platform and to your website's CMS or form-building tool, since the integration touches both sides of the connection.
Give the vendor access to any ad platforms with native lead forms (Meta, Google's lead form extensions) if those are in scope, since these require their own specific API connections separate from your website's forms.
Supply a list of what fields you actually need captured for each lead, name, contact details, the specific service or product they are interested in, their marketing source, so the vendor builds the fields your team will actually use rather than a generic default set.
Ownership and exit clauses
State that your CRM account and all lead data inside it belong to you outright, and that this must be the case regardless of which platform you use, since some platforms make data export easier than others and the vendor should confirm the export process works before you commit.
State that any custom fields, automations, or workflows built inside the CRM are documented in plain language, and that the underlying logic (why a lead gets scored a certain way, what triggers a specific follow-up) is explained, not left as an opaque black box only the vendor understands.
Add a clause requiring the vendor to provide a full data export in a standard format (a spreadsheet, at minimum) on request, and to confirm whether any automations or scoring logic can be exported or only rebuilt from scratch if you switch platforms later.
Milestones and acceptance tests
Set a mapping milestone: a written document showing every lead source, every field it captures, and where that data lands in the CRM, approved before the integration is built.
Set an integration milestone next, proven with a real submission: a real test lead sent through each connected source (a test form, a test call if call tracking is involved) appears correctly in the CRM with accurate source attribution, within a reasonable time window.
Set a data quality milestone roughly 30 days after go-live, checking for duplicate lead records, missing source attribution, or any leads that failed to sync, since integration issues involving specific edge cases (an unusual form field, a slow API response) sometimes only surface under real traffic volume.
Scoring rubric
Weight data portability heavily here: 30 percent goes to the vendor's written data mapping and testing plan, 25 percent to price, 20 percent to experience with your specific CRM platform, 15 percent to data export and portability terms, and the remaining 10 percent to references you actually call.
Score data portability explicitly: ask the vendor to confirm, in writing, that a full data export is possible on your platform, and score down any vendor who is vague about what happens to your automations and lead scoring logic if you switch platforms later.
Ask references specifically whether they experienced any duplicate lead issues after go-live and how quickly the vendor resolved them, since this is a realistic test of ongoing support quality, not just initial setup competence.
Questions every bidder must answer
Make every bidder answer in writing: Which lead sources will be connected, and how is source attribution captured for each one? How do you prevent or clean up duplicate lead records? Can I fully export my lead data and automation logic if I switch CRM platforms later?
Also ask: How do you test that the integration is working correctly before calling it complete? Will I have full admin ownership of the CRM account? What happens to my lead data, fields, and automations the day this relationship ends?
A bidder who can describe deduplication and data export specifically, by platform, by field, is showing real experience with the CRM you are using. A bidder who answers only in general terms about connecting your leads is likely building something narrower than what you actually need.
Related questions
It depends on your budget, your sales process complexity, and whether you need industry-specific features. A licensed platform is faster to launch; a custom CRM can fit your exact process but depends on ongoing developer support to maintain and change.
This requires the CRM integration to be built with automation in mind from the start, a form submission triggering a specific workflow, not just a data connection. Confirm this is explicitly in scope, since it is a larger project than data capture alone.
It depends on how portable the platform's data export is and whether your automations were documented in a way that can be rebuilt elsewhere. Ask this question before you commit to a platform, not after you have built months of history in it.
It works best once you have enough historical data to base the scoring criteria on. A brand-new CRM with no lead history yet may benefit more from getting clean data capture right first, then adding scoring once there is real history to learn from.
Yes. CRM lead-source attribution and pipeline setup are folded into our website and Google Ads engagements, and the written field map plus 30-day data quality review this pack asks for are what we would hand over.
SearchPod is a vendor for this kind of work and would answer this RFP; the acceptance tests here are the ones we agree to.
Want this scoped and priced for your business?
Get a free, no-obligation proposal within one business day. We look at your site and your market and tell you plainly what we would do, and what we would not.
Get your free proposal