Work

What we built,
and what it changed.

Two builds in detail — the problem, the approach we took, and the number that moved.

What changed

+76% first-hour amplification on the ISPOR 2026 campaign
7,800 records enriched into a clean CRM segment, zero errors
Connected systems network funnel

Case 01

The Salesforce-to-growth-engine build

ProblemTheir Salesforce instance wasn't a source of truth for anything. Adoption was low, account ownership had stopped being maintained, flows were broken and Einstein was only half turned on. A test against the contact file projected a 61 percent bounce rate. The ISPOR US conference, the largest deal-generating event on their calendar, was three weeks away.

ApproachWe evaluated the whole org before touching anything, then used third-party data to cure and update contact and account records. From there we built the integrations and agents that connect what sales touches every day: email, the CRMs, event invitations and Slack. Clean information now arrives in Salesforce without anyone retyping it.

Result +76% first-hour amplification on the ISPOR 2026 campaign Record turnout at the event. An 87 percent reduction in typical bounce rate. Every contact activity mapped to Salesforce.

Case 02

The market signal system

Our client puts artificial intelligence at the core of its healthcare products. We built the AI that runs its market signal intelligence—a job that took marketing judgment and engineering depth from the same team.

10 sec reviewer time per signal, down from roughly two minutes, with duplicates reaching the CRM cut by 95 percent

ProblemMarket intelligence arrived as noise. Alerts, intent data, RSS feeds and trade press—hundreds of items a week, each needing a person to decide whether it mattered, whether it was already seen, and which account it belonged to. Roughly two minutes an item, and it never finished. Nothing accumulated.

ApproachA pipeline that triages before a human sees anything. A language model separates noise from signal, collapses duplicate coverage to one canonical source ranked by domain authority, and matches what survives to the right company record. The reviewer decides one thing: standard or priority. Every decision becomes labeled data, so accuracy improves, and review volume falls as a by-product of use.

ResultSignals reach the correct company record in seconds, precision holds above 95 percent, and the CRM stopped being a contact database. It became an accumulating, human-verified history of what has happened to every client, prospect, and competitor account—the kind of asset a rival cannot buy, only build.

A company that could have built this itself chose not to.

Market signals from alerts, intent data and trade press converging into a single organized CRM record

Range

Healthcare is the hero. It isn't the whole story.

 

SaaS

A SaaS platform with plenty of top-of-funnel activity and no honest read on what converted. We instrumented TOFU-to-revenue reporting so the team could finally see what was working.

 

FinTech

A FinTech platform whose team was buried in manual campaign upkeep. We built agents to take the repetitive work off their plates so marketers could spend their time on judgment calls, not data entry.

 

Retail

A retail brand running campaigns on top of disconnected tools. We unified data, hygiene, and attribution across the stack so the engine had clean fuel to run on.
AESARA

“They didn’t just run our campaigns — they rebuilt the system underneath them so we could finally trust the numbers.”

Draft — testimonial pending client approval

Arine

“Pertinacity is the rare team that’s as fluent in marketing judgment as it is in the AI it builds to support it.”

Draft — testimonial pending client approval