Uptoo
How Uptoo Turned a 100K Contact CRM Into a Sales Machine With Cashmyrr
Uptoo's sales reps were drowning in bad data. Cashmyrr cleaned 100K records, enriched 60,000 companies, and embedded one-click enrichment buttons directly inside HubSpot in four months.
130K
contacts audited
45K
records removed or merged
60K
companies enriched
Uptoo Had 20 Years of Sales Data and No Way to Use Most of It
Uptoo is the French specialist sales agency. For over 20 years, they've helped more than 7,500 B2B companies exceed their commercial targets through recruiting, training, management, and tooling for sales teams. Their buyers are CEOs, Sales Directors, and HR Directors, decision-makers who expect outreach to be sharp, relevant, and precise.
Behind that outreach sat a HubSpot CRM telling a different story. Nearly 100K contacts, 60K companies, and 150K deals had accumulated over years with no systematic cleaning process. Much of it was noise: bounced emails, contacts without phone numbers, companies with no associated people, deals that hadn't moved in over a year. The data was technically there. Practically, it was getting in the way.
Two stakeholders on Uptoo's side drove the project: Flavien Candela, Chief Marketing Officer and owner of the pipeline and marketing strategy, and Mathieu Yon, Manager of Sales Performance, responsible for sales process optimisation and the daily tooling of the commercial team.
Sales Reps Were Spending Their Time on the Wrong Tasks
Uptoo didn't call Cashmyrr on a hunch. They measured the problem first. During the summer, leadership ran an internal survey asking sales reps which tasks consumed the most time and which created the least value. The findings were unambiguous and pointed directly at CRM data quality.
"The number of appointments booked by the sales team had stagnated. It turned out that the time spent on qualification in the CRM was too high, searching for phone numbers, enrichment, cleanup, deduplication. There was a direct impact on sales execution.”
- Flavien Candela, CMO, Uptoo
Three pain points dominated:
No reachable phone numbers. A significant share of contacts had no mobile or direct line. Cold outreach over the phone, Uptoo's primary prospecting channel, was slow, hit-or-miss, and demoralizing for reps.
Obsolete and incomplete records everywhere. Companies with no contacts attached, contacts with no email or LinkedIn, deals with no activity in over a year, all of it sitting in HubSpot, degrading every list, every sequence, and every report built on top of it.
No enrichment process inside the CRM. When a rep needed a contact's direct number or wanted to find decision-makers at a target account, there was no standard workflow. Each rep handled it differently, manually, with third-party tools outside the CRM.
The commercial impact was tangible. As Mathieu Yon put it: the problem was "principally a loss of time for the teams on their daily tasks, we lose productivity."
Four Targets Cashmyrr and Uptoo Agreed On
Remove CRM dead weight: contacts, companies, and deals that serve no commercial purpose without touching a single active account, customer, or live deal.
Increase data completeness on key company fields: phone number, LinkedIn, city, region, and industry.
Give every sales rep a one-click phone enrichment button: directly inside HubSpot contact records.
Give account managers a one-click button: to surface missing decision-makers at target companies, without leaving HubSpot.
The overarching goal was clear: free up time that sales reps were spending on low-value CRM tasks and reinvest that time into prospecting. As Mathieu framed it, the objective was "to increase the number of meetings booked per rep by offloading time spent on tasks with low added value."
Cashmyrr's Approach: Clean the Foundation, Then Build on Top of It
Cashmyrr structured the engagement around two parallel workstreams. The first was a one-shot, high-stakes data cleaning operation across the three core HubSpot objects: contacts, companies, and deals. The second was the design and deployment of two enrichment buttons embedded directly inside HubSpot, connected to external data providers through automated workflows.
The logic was deliberate: there is no point building enrichment on top of a dirty base. An accurate phone number added to a duplicate contact, or a freshly retrieved decision-maker attached to an obsolete company record, creates more confusion than it solves. We cleaned first. Then enriched.
Uptoo also had a specific expectation around tooling: they wanted guidance on which enrichment providers to use. We ran a structured evaluation and recommended the stack best suited to Uptoo's use case, balancing coverage, cost, and data freshness.
How Cashmyrr Executed the Project
Auditing 365K CRM Records to Understand the Real State of the Data
Before touching a single record, Cashmyrr ran a full audit of Uptoo's HubSpot instance. The scope was substantial: 100K contacts (each with up to 324 properties), 60K companies (224 properties), and 150K deals (369 properties).
The team mapped each object's key identifiers: email, phone, LinkedIn, domain, using a priority waterfall approach. In plain terms: for each field, we checked multiple HubSpot properties in order of reliability and kept the first non-null value. For a contact's country, that meant checking the country field first, then falling back to ip_country. This produced a normalised view of each record's actual data quality before any decision was made.
Cashmyrr then defined inactivity thresholds: 360 days for contacts and companies, 60 days for deals and classified each object: Customer, Linked to a Deal, Has a Scheduled Next Step. These three categories were fully protected from any cleaning action, regardless of other signals.
Every Deletion Followed Explicit, Traceable Rules: No Guesswork
Every deletion and merge followed versioned rules. Nothing was heuristic; nothing was manual.
Contacts were flagged for removal if they matched any of the following conditions (and were not protected by an exclusion rule): hard-bounced email, no identifier present (no email, no phone, no LinkedIn), no association with any company or deal, inactive for more than 360 days, field completion rate below 30%, or a generic email address (e.g., team@, support@).
For duplicate detection, we used three merge keys: matching LinkedIn profile, matching email, and a name-based match, first name and last name, lowercased, stripped of spaces, combined with a shared company association. That combination is a strong signal that two records represent the same individual.
Companies followed similar logic: removal for records with no contacts, no deals, low field completion, or extended inactivity. Duplicates were merged by LinkedIn page, domain, or name (with an additional check to exclude companies that were intentionally associated with one another, for example, subsidiaries).
Deals were archived if they were stale: open or ambiguously closed, no scheduled next step, fewer than 10 logged sales activities, and inactive for 60+ days. Deals explicitly marked as won or lost with a recorded close reason were never touched.
The Cleaning Impact: Before and After
Object | Before | After cleaning | After merging |
|---|---|---|---|
Contacts | 130K | 92K | 92K |
Companies | 70K | 42K | 41K |
Deals | 160K | 100K | 100K |
In total, Cashmyrr removed or merged approximately 40K contacts, 30K companies, and 60K deals, all of them dead weight that was degrading list quality and prospecting efficiency. Not a single active account, customer, or deal with a scheduled next step was affected.
Enriching 60K Companies in One Pass
With the base cleaned, Cashmyrr ran a bulk enrichment operation over the company database. Out of ~70K companies in scope, 60K were successfully enriched, matched against external providers to fill gaps across LinkedIn, phone, address, city, region, country, industry, headcount, and company description.
The ~5,000 companies that returned no match were largely those with no domain on record, making a reliable external match impossible.
Company data completeness before and after Cashmyrr's enrichment:
Field | Before | After |
|---|---|---|
80% | 88% | |
Phone (generic) | 80% | 91% |
Address | 79% | 91% |
City | 71% | 85% |
Region | 64% | 83% |
Country | 86% | 92% |
Employees (headcount) | 81% | 86% |
Company description | 91% | 94% |
Two One-Click Enrichment Buttons, Shipped Directly Inside HubSpot
The enrichment buttons were the second major deliverable and the one Uptoo's team valued the most in daily use.
Cashmyrr deployed two buttons embedded directly in HubSpot record views, each triggering n8n-orchestrated workflows. n8n is an automation platform that connects different apps and APIs through visual workflows think of it as the invisible engine running behind the button.
"Enrich Phone" button on contact records.
Enrich Phone
A sales rep clicks one button. Behind the scenes, the workflow queries a cascading stack of three enrichment providers, Kaspr first (synchronous, meaning the result comes back immediately), then FullEnrich (asynchronous, with a webhook callback meaning the system waits for the provider to respond later), then Zeliq as a final fallback. As soon as a valid direct dial or mobile number is found, it's written back to the contact's mobilephone field in HubSpot. A custom tracking event logs whether a number was found, from which provider, and how many results came back. The previous phone value is preserved in property history, so nothing is ever lost.
This multi-provider cascade was a deliberate architecture choice. No single enrichment provider covers every contact. By chaining three in priority order, each with different strengths in coverage and geography, the system maximises the chance of returning a usable number on any given click.
When asked which single element of the project brought the most value, Flavien was direct: "The phone enrichment button on contacts."
"Get Contacts" button on company records.
Get Contact
From any company record, clicking this button triggers a workflow that searches for decision-makers at that account. The system queries Apollo's database with 30+ targeted job titles (CEO, Sales Director, HR Director, and equivalents), filters results through a GPT-based relevance check to eliminate false positives, enriches emails, assigns each contact a standardised job function, and creates the new contact records directly in HubSpot associated with the right company.
This compressed what was previously a multi-step, multi-tool process (LinkedIn search → contact data tool → manual entry into the CRM) into a single click that produces structured, associated, ready-to-prospect records.
Automated Monthly Cleaning to Prevent Data Degradation
We also deployed a scheduled n8n workflow that runs monthly. It automatically identifies records that have fallen into pre-defined cleanup lists (bounced contacts, orphaned companies, stale deals) and removes them keeping the database lean without requiring manual intervention from Uptoo's team. This was a critical piece: without ongoing maintenance, a cleaned CRM starts degrading again within weeks as new data enters the system.
Full Technical Documentation
We delivered a complete documentation package covering the architecture of both enrichment buttons, the n8n workflow logic for each step, the enrichment provider configuration, the cleaning rules, and a maintenance runbook. Uptoo's technical team can understand what runs when, trace any decision, and troubleshoot independently.
Results
What Uptoo Walked Away With
A clean CRM foundation: 45K contacts, 30K companies, and 50K deals removed or merged across all three objects, following explicit, auditable rules that protected every active account, customer, and live deal.
60K companies enriched in a one-shot bulk operation, with measurable field completion gains across 10 key properties including phone coverage jumping from 80% to 91% and region data from 64% to 83%.
"Enrich Phone" button deployed in HubSpot: one click to retrieve a direct phone number for any contact, powered by a three-provider cascade (Kaspr → FullEnrich → Zeliq) with full tracking and property history preservation.
"Get Contacts" button deployed in HubSpot: one click to surface and create decision-maker contacts at any target company, with GPT-based filtering and automatic job function assignment.
Automated monthly cleaning workflow running on n8n to prevent data degradation over time.
Full technical documentation covering workflows, provider configuration, cleaning logic, and a maintenance runbook.
From Noise to a CRM the Sales Team Actually Uses
The most direct outcome was a HubSpot instance that Uptoo's sales team can trust. Before Cashmyrr's engagement, running a prospecting sequence on a contact list meant wading through bounced emails, missing phone numbers, and companies that no longer existed. After the cleaning and enrichment, the base is scoped to records that meet a clear quality threshold, associated, identifiable, and active.
The gains on company enrichment are concrete and directly affect daily prospecting. Phone coverage on company accounts went from 80% to 91%. City-level location data, critical for territory-based sales, went from 71% to 85%. Region went from 64% to 83%. These aren't cosmetic improvements; they determine which records a rep can actually call, and how accurately they can segment their territory.
The enrichment buttons changed the daily workflow. Where a sales rep previously spent several minutes per contact searching for a direct dial across external tools, the phone enrichment button collapses that to a single click with the result written directly into the CRM record. Uptoo's internal feedback confirmed: the phone enrichment button was the single most valued deliverable of the engagement.
It's worth being transparent about the full picture. While the project delivered measurable time savings on daily CRM tasks, qualification, enrichment, number lookup, Mathieu Yon noted that the time freed up has not yet fully translated into more meetings booked: "the time saved isn't necessarily reinvested in prospecting that's a separate challenge." This is a common pattern in RevOps projects: operational gains are immediate, but behavioural change in how reps allocate their time takes longer and requires management reinforcement beyond the tooling layer.
"It saves time and brings expertise to companies that may be lacking it.”
- Mathieu Yon, Manager Sales Performance, Uptoo
What This Project Unlocked
The four-month engagement gave Uptoo's sales team a CRM they can prospect from with confidence. The combination of a cleaned base and embedded enrichment tools means reps spend less time qualifying and more time reaching the right people.
But the project also laid the groundwork for what came next: a Sales AI Agent connected to HubSpot that uses this same clean, enriched data as its source of truth. Without Cashmyrr's cleaning work, the agent would have been reasoning on noise. The data quality project was, in that sense, not just a maintenance task, it was the strategic foundation for Uptoo's entire AI-assisted sales stack.
