/w=256,quality=90,fit=scale-down)
Jacob TuwinerEvery sales leader knows the frustration of unreliable CRM data—duplicate entries, outdated information, and questionable accuracy turning even the most advanced sales teams into glorified data-entry clerks.
ConnectWise faced exactly this issue, dealing with nearly 200,000 messy prospect records and an inability to confidently identify genuine Managed Service Providers (MSPs) in their SMB-focused target market.
Traditional methods like legacy data vendors and manual offshore teams provided lackluster results, slow timelines, and disappointing accuracy.
ConnectWise needed a fast, scalable, and cost-effective solution that could deliver highly accurate account enrichment at scale.
That's when Rick Collins, VP of Demand at ConnectWise, turned to Sculpted and Clay, unlocking a new standard for their CRM data hygiene cleanup which revolutionized their sales team's efficiency.
TL;DR
Challenges | Solutions |
❌ Data Quality & Duplication: Untrustworthy data across ~200,000 prospect accounts. | ✅ Leveraged Clay integrations & custom AI prompts for flexible, accurate data enrichment at scale. |
❌ Identifying True MSPs: Could not reliably separate real qualified accounts from spam. | ✅ Created custom AI prompts to accurately categorize accounts into 5 tailored industry segments. |
❌ Manual Website Verification: Needed manual website checks to validate TAM fit. | ✅ Automated website scraping with Claygent AI, drastically reducing manual effort and cost. |
❌ Lost Sales Productivity: Sales team forced into manual data entry instead of selling. | ✅ Enriched and scored entire database automatically, freeing sales reps to focus solely on selling. |
❌ Legacy Data Enrichment Limitations: ZoomInfo provided only 50-60% accurate enrichment. Equally bad results with offshore manual enrichment services. | ✅ Achieved 95%+ accuracy with nuanced, custom AI prompts surpassing legacy vendor limitations. |
The Challenge: A CRM Data Dumpster Fire
As Rick puts it, every CRM faces data quality and duplication challenges. In fact, he's never encountered anyone who says "I love my CRM data."
"I would say all CRMs have a duplicate challenge and a data quality challenge. I've never met anybody that said their CRM was really clean.”
- Rick Collins, VP Demand @ ConnectWise
For ConnectWise, this challenge manifested in roughly 200,000 prospect account records filled with unreliable data. The team struggled to differentiate between spam accounts and legitimate, active Managed Service Providers (MSPs).
Since their target market consists of SMB-focused MSPs, traditional classification methods fell short—there's no standardized SIC code or industry category to easily identify qualified accounts in platforms like ZoomInfo or Sales Navigator.
Instead, team members had to manually visit each company's website to verify whether they offered IT services or managed services, determining if the prospect truly fit their TAM.
As a result, their elite sales team was reduced to data entry specialists, spending precious time combing through CRM records instead of selling.
Before discovering Clay + Sculpted, ConnectWise had experimented with several conventional solutions—all of which delivered disappointing results.
Usual suspect #1: Legacy data vendors like ZoomInfo
Although they do a good job of giving a high-level view of account and contact data, ZoomInfo was unable to answer questions like:
- Is this company still in business?
- Are they located in a serviceable geography?
- Most important of all, are they truly an MSP or not?
While larger companies above 20 employees had okay coverage, the coverage for smaller companies (under 5-10 employees and the bulk of the market) was much less. Across the board, ZoomInfo enrichment was only 50-60% accurate at best.
Usual suspect #2: Offshore teams
After data enrichment with ZoomInfo failed, ConnectWise turned to offshore teams in the Philippines to manually review and enrich every record.
The researchers combed through tens of thousands of records, checking if the website was live, where they were located, employee count, and classifying the company as an MSP or another type of company.
Still, after the slow, expensive, manual verification process, the data was still 60% accurate at best. After investing so much manual labor, the data should’ve been 100% accurate.
"We used an offshore team to manually validate, but even then, we only reached about 60 to 70 percent accuracy—it just wasn't good enough. It took us seven or eight months to get through roughly 75,000 accounts, which was a huge investment of time when we needed to generate pipeline quickly. I would manually review a sampling of 50 accounts and consistently say, ‘Guys, six out of 10 isn’t good enough—if we're doing manual validation, it needs to be 10 out of 10.’ Ultimately, there just wasn't a good solution available for a massive database cleanup like this."
- Rick Collins, VP Demand @ ConnectWise
Worse still, it took 7-8 months of time to work through roughly 75,000 accounts, and only with 60% accuracy.
The Solution: Data enrichment with Clay + Sculpted
Clay proved ideal due to its extensive integration options, providing the flexibility needed to customize the enrichment model for ConnectWise's ICP.
Two features were essential to the cleanup project: Claygent (Clay's native AI web scraper) and their ChatGPT integration.
Rather than manually reviewing websites for "Managed Services" or "IT Support" keywords, Claygent could automatically scrape websites at a minimal cost.
Using Claygent, we generated summaries of each company's activities and services, then fed these findings into an AI prompt that categorized accounts into ConnectWise's 5 primary company segments.
By crafting our own prompts, we captured nuances that ZoomInfo's standard industry filters would miss—even if they supported the MSP SIC code. This approach achieved unprecedented 95%+ accuracy compared to legacy enrichment tools.
The cost efficiency was remarkable: each record cost mere fractions of a penny using the OpenAI API, compared to expensive human labor. The result was superior accuracy and coverage, delivered faster and more economically than traditional methods.
Why clean CRM data with Clay vs another automated outbound solution?
Rick explored several AI platforms before choosing Clay. One platform offered to process 1,000 accounts with deep research and automated outbound campaigns. However, Rick rejected this approach for three key reasons:
Too expensive
The alternative tool was prohibitively expensive for just 1,000 accounts. With ConnectWise's database of 200,000 records, automated outbound platforms weren't financially viable.
Clay, however, could enrich their entire database with superior accuracy and coverage compared to legacy vendors or offshore teams—all for just pennies per record.
But even if Rick had found a more affordable automated solution, another crucial factor influenced his decision:
Enabling the existing team of 39 reps
ConnectWise's large sales team needed quality accounts to pursue. Rick had to prioritize supporting his existing team rather than implementing an automated outbound solution that would sideline his reps.
While the reps handled their own research and prospecting, they relied on accurate account data to meet pipeline quotas. They couldn't be left without support, making Clay's CRM database cleanup the clear choice.
Deliverability
Finally, email deliverability posed a major challenge. With increasingly difficult inbox placement in 2025 and beyond, automated outbound solutions are becoming harder to justify.
No matter how well-researched your list or how personalized your emails, they're worthless if they never reach the inbox. You could send tens of thousands of emails without any response.
In contrast, a sales rep sending targeted, personalized emails and making live calls to perfect-fit prospects remains a timeless strategy. The challenge used to be finding those qualified accounts—but now Clay automates that process, letting the sales team focus on what they do best: sell!
Sculpted’s CRM cleanup process
1. Scoping
“We quickly honed in on the pain we wanted to solve, how we were going to do it, and exactly what we wanted to enrich—deciding which Clay tools and connectors we’d use, and clearly defining how many records were involved. Sculpted guided us through developing the categorization model, determining what made an account a good fit versus not. The iterative process was really valuable; we'd review samples, tweak the model, and significantly improved accuracy through multiple iterations. Sculpted also proactively managed our resources, frequently checking in to confirm accuracy before burning through credits, ensuring efficiency and cost-effectiveness throughout.”
- Rick Collins, VP Demand @ ConnectWise
Sculpted starts by understanding your Ideal Customer Profile (ICP) in depth. We combine this with our expertise in Clay and other systems to surface insights that qualify accounts accurately, consistently, and cost-effectively.
During the kickoff call, we explore your ICP by asking: "What does your dream account look like and why?"
We then map these answers to specific CRM properties that will store the enrichment data.
After that, we create a property enrichment plan for each field, determining the best data source—whether that's Clay integration, AI, or external APIs.
The industry category definitions require particular attention. We develop detailed rules to ensure the inputs generate accurate classifications.
The industry definition table below demonstrates how precise and customized our category and subcategory definitions can be:
2. Calibrating the enrichment and scoring model
After scoping the enrichment model, we requested a sample dataset of 20-30 records to fine-tune its accuracy.
We built the enrichment workflow, tested it on the sample data, and refined it based on ConnectWise's feedback. Through this quick, iterative process, we achieved 95%+ accuracy for both account and contact records.
“The iterative process with Sculpted was great—we'd quickly review samples, tweak the model, and rapidly improve. Within just a few iterations, we got exactly where we needed to be.”
- Rick Collins, VP Demand @ ConnectWise
We then developed a custom scoring model that considered factors like industry, employee count, and location to assign accounts into tiers 1-3.
When Rick presented the results to the SDR leadership team for critique, their response was overwhelmingly positive: "No, this is really good... if we can get more data like this, we'll be in great shape."
3. Full database enrichment and CRM import
After the model is honed in, we enriched their entire dataset, scored the accounts and contacts to surface the most qualified records, and pushed the data back into their CRM for the sales team to action.
“Once the model was solid, Sculpted quickly turned around high-quality data. I even asked our SDR managers, ‘Poke holes in this, tell me what's wrong,’ but they came back saying, ‘No, this is solid—if we get data like this, we're going to be in really good shape.’ The speed at which Sculpted delivered such accurate data was amazing, especially after we'd previously spent months manually validating with limited success.”
- Rick Collins, VP Demand @ ConnectWise
Now, Rick’s team focuses on prospecting into Tier 1 Accounts with Tier 1 decision maker contacts at those accounts.
Why ConnectWise hired Sculpted to implement Clay
Rick hired Sculpted for two reasons: Sculpted’s deep understanding of the CRM data problem, and accelerated time to value.
Sculpted’s deep CRM & Clay expertise
First of all, Rick felt Sculpted understood their CRM problem better than any other agency during the evaluation process.
"It took me a minute of explaining the problem and it was like, 'Yeah, we can do this, this, this.' It was just a clear understanding of the pain that I was going through and how you could solve it."
- Rick Collins, VP Demand @ ConnectWise
Other “Claygencies” are generally less technical, and more focused on outbound automation than CRM data hygiene. Rick felt he was speaking a foreign language with them… which is not what you want when you’re bringing in an agency to help with CRM data hygiene.
Accelerated time to value
“We chose Sculpted because we simply didn't have the bandwidth or months to spend onboarding and training internally. Sculpted gave us the fastest path to value—someone who fully understood our pain and knew exactly how to solve it quickly.”
- Rick Collins, VP Demand @ ConnectWise
Taking on a data project in-house is challenging. Even with substantial internal revenue operations resources, the timeline can stretch for months.
New tool training and constant distractions derail progress. Between board meeting preparations, urgent CRO report requests, and other priorities, projects lose momentum.
Without dedicated resources for list building, account segmentation, and data enrichment, completion becomes nearly impossible. The process of procuring Clay, training staff, and managing competing priorities would have taken 6-8 months or longer in-house. That's why partnering with an agency like Sculpted offered the fastest path to value.
What’s next
In phase 1, we cleaned up ConnectWise's 200,000 prospect account database to identify their core market of tier 1 MSPs.
They also have 150,000 unconverted leads sitting dormant in their database. We'll use Clay to enrich these records while also analyzing the non-MSP accounts.
Large companies with internal IT teams, for instance, are prime candidates for ConnectWise. We're developing a new scoring model to categorize these companies into six industry segments based on ConnectWise's ideal customer profile. This will help them extract more value from their existing database.
Finally, we'll apply this scoring model to new records before they enter the CRM. This will help ensure higher quality data when importing lists from ZoomInfo and other sources that previously introduced low-quality records.