Case Study

Assembled: From Attribution Guesswork to a 35% Pipeline Surge

In discussion with

Lindsey Marymont
Assembled

About

Assembled is the AI platform for customer support providing autonomous AI agents and AI-powered workforce management built for modern customer support teams. Founded in 2018, Assembled is the support automation for teams at Autodesk, Canva, Carta, Clear and Hubspot and is the only platform combining the power of AI agents with human workforce management.

Industry

Software Development / Customer Support Operations

Company size

51–200 employees

Headquarters

San Francisco, California

Founded

2018

GTM Team

9 SDRs, 7 marketers, 13 AEs across enterprise and mid-market

Lindsey Marymont
“I've been in marketing for 10+ years, and I've never been able to get a clear picture of the buyer journey. But with Upside, we are able to see a complete milestone analysis with data pulled from Gong we wouldn't ever be able to see in Salesforce.”
Assembled

Lindsey Marymont

Head of Demand Generation

Metrics

The Data Hidden in Plain Sight

30k+ Hidden Contacts Discovered

Upside found over 30,000 people (~14% of total contacts) who joined Assembled's calls and influenced buying decisions but never appeared in Salesforce or HubSpot.

37k+ Hidden Touchpoints Uncovered

Upside revealed over 37,000 interactions from Gong recordings, demos, discovery calls, and check-ins that were never logged as activities in any CRM.

2x More Accurate Pipeline Data via MCP

When Assembled connected Upside's MCP to their AI tool (Dust), pipeline queries returned roughly twice the completeness and accuracy compared to querying Salesforce directly.

Highlighted Results

Upside gave Assembled data-driven clarity that reshaped channel strategy, redirected budget, and surfaced a hidden growth engine.

  • True Event Impact Uncovered: Upside confirmed that events drove ~20x more pipeline per opportunity than other sources such as paid advertising. Lindsey used this data to repurpose ad strategies as amplification and invest heavily in events — a decision she couldn't have justified without it.
  • AEO Win Rate Lift of +12.6 Percentage Points: Buyers who discovered Assembled through AI engines like ChatGPT and Claude closed at a measurably higher rate. After months of investing in AEO efforts, Upside was the only tool that could surface this signal and validate this investment.
  • Full Buyer Journey Visible for the First Time: Milestone analysis pulled Gong call data, meeting touchpoints, and email threads into a single view, replacing hours of manual attribution done in spreadsheets.
  • Executive Use Case: The exec team reviews pipeline weekly and relies on Upside data to reveal opportunity analysis and deal sourcing.
  • Immediate MCP Adoption Creates Team Alignment: With over 2,300 MCP calls in just the first week, Assembled's AI workflows are fueling the team to have conversations about what the data is revealing instead of spending time validating the data.
Background

A Product Built on Support-First Principles

Assembled has been powering the highest caliber support teams since its inception in 2018. Starting with core workforce management, they found a true need for modern solutions to legacy problems — helping teams with scheduling agents, forecasting support ticket volume, and providing real-time adjustments for things like agents calling out sick or surging support requests from a product bug or meme stock virality (that really happened).

As the first true omni-channel workforce provider, teams running Assembled could manage support across phone, chat, email, and emerging channels like social and text. Later Assembled built solutions for outsourced support teams, providing solutions like Vendor Management so customers could manage in-house support as well as productivity and volume of outsourced call centers.

Today, those innovative support-first principles have led to the only solution on the market that helps CX leaders manage human agents, AI agents, and outsourced BPOs to provide the best support possible. For many teams this means a small support team of 10 is able to do what would require 50 or more agents previously.

With a GTM team spanning 9 SDRs, 7 marketers, and 13 AEs across enterprise and mid-market, Assembled is at the forefront of the AI revolution. They compete against native AI-agent players on one side and legacy workforce management incumbents on the other. They're the only solution that combines both.

The Challenge

Assembled's Attribution Draws Misleading Conclusions

As Assembled transitioned from a strictly software business to a multi-product AI offering, they needed a way to understand how to land and grow customers for their emerging tech. Many potential customers were quick to buy into the latest AI support provider but didn't launch formal evaluations. Or equally common, teams were evaluating AI for support and were finding options in unconventional ways. Assembled needed to understand the new way of buying and stay top of mind for CX leaders.

"The challenge we were trying to solve was essentially annual planning, getting a good sense of where our pipeline was coming from, where to double down as we were thinking about budget allocation for the year ahead."

The team's tech stack included 6sense, Salesforce, HubSpot, Gong, and Default among others, but none of these tools talked to each other meaningfully. When a big deal came in, figuring out how it started meant manually piecing together call transcripts, Gong recordings, Salesforce fields, and email threads.

"I had to piece together Gong calls, transcripts, Salesforce … and most of it just lived in the rep's head. I was constantly asking, 'hey, how did this come to us?' as we were uncovering buying patterns for a new product."

Their first-touch attribution model, run through Salesforce, had fundamental limitations: it only captured the first marketing touch, couldn't connect Gong call data to deals, and left entire channels like events and AI discovery unmeasured.

Finding the Right Partner

Before choosing Upside, Assembled also evaluated other solutions such as HockeyStack.

"HockeyStack had the visualization but not the data. They weren't willing to work with us throughout the evaluation, they just did demos. Upside did a full pilot with our actual data. And the setup was made so easy for us. Data integrity was the most important thing, and we needed to validate that above everything else."
The Outcome

Multi-Touch Attribution That Actually Works with Imperfect Data

After careful evaluation, Assembled chose Upside.

"The fact that Upside could work with imperfect data was the selling point. We knew our data wasn't clean. Getting all the Gong transcripts and everything else into a single milestone view — that was something I've never seen before."

With Upside, Assembled can now see the complete influence picture across 16 distinct channels and 4.9 million touchpoints, with no manual exports and no battles with reps to log activities.

"Instead of nagging reps to log things in Salesforce and connect their email, Upside just did it for us."

Aha Moment #1: The Buyer Journey, Finally Visible

The milestone analysis was the feature that changed everything for Lindsey.

“I've been in marketing for 10+ years, and I've never been able to get a clear picture of the buyer journey. I've spent entire quarters trying to export data and build it in spreadsheets, but all you get is campaign data. With Upside, we got a milestone analysis that pulled in Gong data and surfaced things that weren't in Salesforce.”

Aha Moment #2: Activation After the Event

This was Lindsey's sleeper favorite — a metric that exists nowhere else.

"It is possible to use datestamps to see when leads came in vs. when they converted, but it's not easy. Especially if you use Salesforce campaigns as a parking lot for list imports — what actually makes leads convert isn't reflected in the data. Being able to see the progression after a campaign touchpoint was a big deal, especially for events and other offline activity sources. That timestamp is nearly impossible to get in a Salesforce report. And this metric doesn't exist anywhere else."

Aha Moment #3: Uncovering the Events Truth

The data revealed something the team had suspected but couldn't prove. Events, despite influencing fewer raw opportunities than email or meetings, drove the highest pipeline of any channel.

"I had the data to make the case: double down on events. We were meeting all the right people and able to engage them live with a demo and discovery of their specific use cases. These conversations lasted well beyond the event and often led to intros with buyers. With Upside we could see the impact these conversations were having on our bottom line and made the case to invest more in events and even hire a second events marketer."

Discovering a Hidden Channel: AI Engine Optimization (AEO)

One of the most novel findings: buyers were using AI chatbots to discover and evaluate Assembled but it was extremely hard to quantify. With Upside, they proved the value. A study across 521 opportunities, combining Gong call recordings, UTM citation traffic, and Salesforce notes revealed an AEO attribution goldmine:

AEO Discovery Method

Call Mentions78%Gong recordings withAI tool mentionsCitation Traffic22%LLM UTM / referral0%overlap

No overlap between discovery methods — call mentions and citation traffic were entirely separate paths.

The win rate told the real story: AEO-influenced opportunities closed at a 12.6 percentage point higher rate than non-AEO opportunities. Buyers who used AI to discover Assembled were more qualified and more likely to buy.

"Honestly, I found you by going on ChatGPT. I was like, 'what's the best WFM platform?' And then you guys came up and that's how I found you guys." — Assembled prospect
"As AI-driven discovery started to pick up, Upside helped us see it on the back end — being able to actually track that was really exciting."

What Buyers Were Searching For Using AI Tools

% of 22 AEO-influenced opportunities · Buyers often searched for multiple themes

WFM Tool Discovery77%(17)Vendor Comparison50%(11)Integration Research27%(6)Capability Matching23%(5)Pricing Research9%(2)

From Dashboard to AI-Native Workflow: The MCP Story

Before Upside, Assembled's internal AI tool (Dust) queried Salesforce and HubSpot directly. When they connected the Upside MCP, the impact was immediate: pipeline queries returned roughly 2x the accuracy and completeness compared to Salesforce. In the first week alone, Dust made over 2,300 MCP calls, used by the CEO, marketing team, and ops.

"Things that used to take me hours to pull, I'm getting them in Upside in minutes. Instead of having to pick what questions to ask based on how long it will take to get a verifiable answer, with a quick MCP search, I'm uncovering issues I didn't even know existed."

Today, the executive team runs every major opportunity through Upside each week, checking pipeline trends, deal sources, and channel influence before key decisions.

"The executive team uses Upside every week. We look at pipeline week-over-week, find the major opportunities on everyone's radar, run them through Upside, and figure out how they came to us and what's happening with them."
What's Next

A Partnership That Keeps Delivering

With clear data validating their channel bets and a weekly executive rhythm built around Upside, Assembled is now expanding how they use the platform:

  • Deeper event investment optimization informed by milestone data and post-event activation rates
  • Broader MCP adoption across more of the marketing and ops teams
  • Salesforce automation via connected MCPs to reduce manual data entry
  • AEO content investment guided by the win rate data Upside surfaces
"The partnership matters as much as the product. It wasn't 'the trial's over, bye.' You're still here with us. Hosting events, fostering networking, and genuinely caring about our business — that's hugely important to who you do business with."