Case study · Pardot between ZoomInfo and Sales Cloud
Three five day campaigns in Pardot.
One lead sales would have missed became a $370,000 deal for the client.
A US freight company bought leads in ZoomInfo and sent its campaigns from Salesforce, while the Pardot licence it already paid for sat unused. Working in the team that implemented it, I connected ZoomInfo, Pardot and Sales Cloud into one path, moved the campaigns into Pardot and rebuilt qualification around behaviour, so sales received only the leads that were acting like buyers.
This is for you if you sell to other businesses, buy contact data, already pay for Pardot, and your sales team still calls down the whole list.
From my time in a team that implemented Pardot for clients, remote. The client is not named and no campaign figures are published. The deal amount is shown with the client's result in mind: the sale was theirs, the work made the lead visible. Reviewed 29 October 2026.
What changed
What changed for the client?
Before, a campaign ended when the email left Salesforce, and sales worked whole bought lists by phone. After, every open and click was recorded against the prospect, behaviour decided who was ready, and sales was alerted when a prospect engaged. Three campaigns of five days each were enough for one of those leads to close as a large deal.
3 × 5 days
three campaigns of five days each, sent from Pardot rather than from Salesforce
1 deal
one lead picked out by its behaviour closed as a $370,000 deal. The sale was the client's; what the work changed is that the lead was seen
Every open and click
recorded against the prospect sales works from, where before the send was the last thing anyone knew
0 tools bought
Pardot was already paid for. No new software and no extra lists, the same leads read properly
Before
What was going wrong before?
The money went into data, and the visibility stopped at the send. The client bought good leads in ZoomInfo and emailed them from Salesforce, and after that there was nothing: no record of who opened, who clicked, who was interested. Sales chased cold contacts down broad lists while the people actually reading were buried in them. Pardot, the tool built for exactly this, was licensed and idle.
What was built
Four changes, each one replacing a habit.
One path from ZoomInfo to Sales Cloud.
Leads bought in ZoomInfo, campaigns and tracking in Pardot, the qualified lead and the deal in Sales Cloud, as one flow instead of three separate tools.
Campaigns sent from Pardot.
The campaigns moved from Salesforce sends into Pardot, so every open, click and visit is tracked against the prospect who made it.
Qualified by intent, not by activity.
Lead scoring rebuilt around the behaviour that signals readiness to buy, and only those leads passed to sales.
Sales told when a prospect engages.
A real time alert when a qualified prospect acts, so the call comes while the interest is current.
The decisions that mattered
Which decisions are usually made the other way?
Three choices that are often made the other way. The client's own rule set is not in my files, so the detail below comes from the checklist the team ran its Pardot implementations from, and my own reasoning is marked as such.
Fix visibility before buying more leads.
The usual reaction to a campaign that does not convert is a bigger list or another send. That fails quietly: more cold contacts dilute the few warm ones, and sales stops trusting the list at all. Here nothing was added. Not more emails, not more leads: the same leads went through Pardot, and one of them turned out to be the buyer.
Grade for fit, score for intent, and never let one stand in for the other.
The team graded prospects on four fields, company size, job title, industry and country, and kept engagement as a separate score. My reading of why that matters with bought data: a list bought to match your ideal customer grades well almost by construction, so fit alone cannot rank it. Fit is what ZoomInfo sells; behaviour is what only Pardot records. Sales needs both before a lead is worth a call.
Clear the old scores before the new model goes live.
Scores left from earlier sends make a lead that went quiet months ago look hot today. The team's clean up took them down in thirty day steps by last activity: ten points off at 30 to 60 days, twenty at 61 to 90, and so on to fifty off at 151 to 180 days, with a separate rule beyond that. The rules ran once through the database and were then deleted, so the new model started from current behaviour only.
Lessons
What would I do differently next time?
Looking back at the reporting and lead lifecycle kit the team carried between client orgs, a copy of which sits in my own org and was reviewed again on 26 October 2026, three things.
Stamp the lifecycle dates in the same release as the fields.
The kit had a date field for each lead status and formulas for the hours spent in each, but nothing filled the dates, so the hours stayed empty. Those hours are the measure that proves a story like this one: how long a hot lead waited for sales. Next time one flow on the lead stamps every status change before the first campaign goes out.
No record Id written into an automation.
The flow that creates campaign member statuses started on a record type Id copied from another org, so in a fresh org it never ran, and nothing reported an error. A check on the record type's name instead of its Id travels between orgs.
Test a formula on both outcomes.
The hours in Qualified formula had its condition reversed: empty for leads still open, growing forever for converted ones. One test on an open lead and one on a converted lead would have caught it the day it was written.
What buyers ask first
Three questions before you start
We already pay for ZoomInfo. Why would Pardot change anything?
ZoomInfo tells you who fits. It cannot tell you who opened your email this week and came back to the pricing page, and that is what Pardot records. This client already held the Pardot licence, so the result came from a tool already paid for, not from a new purchase. What each edition costs today is on Pardot pricing in 2026.
Our Pardot was set up years ago and nobody uses it. Do we start again?
Usually not. Here the licence existed and sat idle, and the work was moving campaigns into it, tracking behaviour and rebuilding qualification, not replacing anything. If you are holding back because you heard Pardot is being switched off, it is not, and the status in 2026 is written out with Salesforce's own sources. Owning licences with nothing built is the first case on the implementation page.
Will sales trust the scores?
Only if sales agrees the line. Here sales received sales ready leads and an alert when one engaged, instead of a list. On a rebuild, the qualification threshold is set against your own closed deals and signed by sales before any rule goes live. The reasons a model loses that trust are in why Pardot lead scoring stops working.
For your business
Which parts carry over to your business?
Any company that buys contact data and sends campaigns needs the same three things: campaigns sent from the tool that tracks them, a grade for fit kept apart from a score for intent, and old scores cleared before anyone trusts the new ones. The lists, the industries and the threshold change from company to company; the path from bought lead to closed deal does not.
A quick test before anything else: open your last three campaigns in Salesforce and check whether a single open or click is recorded against the people on them. Then look at whether bought leads are a campaign type of their own; in the team's email dashboard, ZoomInfo list sat as a campaign type next to webinar and trade show, so bought data was always reported separately. If the first check is empty and the second has no answer, sales is calling from a list it cannot read.
Rebuilding scoring and qualification so that sales gets the leads that act like buyers is a fixed price: 45 hours, €2,925, about two weeks, with the threshold tested against your own closed deals and signed by sales.
Serhii Skrypnyk · Senior Salesforce Administrator · 7 Salesforce certifications · on the platform since 2018. Reviewed 29 October 2026.
Read next
The question that usually comes after this one.
Why Pardot Lead Scoring Stops Working
Eight reasons a scoring model stops predicting revenue, from missing decay to grading that describes an old customer, each with the check you can run this afternoon.
Pardot Audit: 27 Checks, or Done at €650
The 27 checks a Pardot audit runs, in six groups, with what a bad answer looks like. Run them yourself, or book one fixed day at €650 with a costed fix list.
Salesforce Lead Management: Seven Layers
Seven layers where leads are lost between a form and a closed deal, what breaks in each, and why fixing the scoring model does not fix the handoff to sales.