Customer Research Platforms: Gong vs. UserTesting vs. Manual Interviews

"We need to scale customer research," my VP Product said. "You can't manually interview every customer. We need platforms that automate this."

He was right that I couldn't interview everyone. I was doing 5-8 customer interviews per month, spending 15-20 hours on research.

"What if we got Gong to automatically analyze all our sales calls?" he suggested. "And UserTesting to run quick surveys at scale?"

Combined cost: $45,000 annually

• Gong: $30K for conversation intelligence
• UserTesting: $15K for rapid user research

The pitch: Automate customer research, get 10x more insights in half the time.

I approved both platforms.

Twelve months later, I'd analyzed thousands of sales calls and run dozens of UserTesting studies. The most valuable insights still came from my manual customer interviews.

Here's why automated research platforms gave me breadth without depth, and why I eventually went back to manual interviews supplemented by lightweight tools.

Platform 1: Gong ($30K)

Sales was already using Gong for call recording and coaching. Product suggested PMM use it for customer research.

The promise: "Analyze every sales call automatically. Surface insights without manual interviews."

Month 1: The Setup

Gong was already recording all sales calls. I got access to the full library:

• 400+ calls per month
• Automatic transcription
• Keyword tracking
• Topic clustering
• Competitor mentions flagged

I set up trackers for:

• Competitor mentions (which competitors come up, what objections)
• Feature requests (what customers ask for)
• Pain points (problems they're trying to solve)
• Use cases (how they plan to use product)

Gong dashboard showed:

• Competitor X mentioned in 45% of calls
• Feature Y requested 23 times
• Pain point Z mentioned 67 times
• Use case A discussed in 89 calls

This looked incredibly valuable. Hundreds of data points without conducting a single interview.

Month 2: The Surface-Level Problem

I started using Gong insights for positioning and messaging.

Insight from Gong: "Competitor X mentioned in 45% of calls"

I clicked through to listen to actual mentions:

• Call 1: Prospect asked "How do you compare to Competitor X?" (generic question, AE gave standard answer)
• Call 2: Prospect said "We're also looking at Competitor X" (just naming alternatives, no detail on why)
• Call 3: AE mentioned Competitor X proactively in pitch (AE bringing it up, not prospect)

45% of calls mentioned Competitor X, but most mentions were surface-level. Not deep competitive analysis.

What I needed: Why prospects preferred us or Competitor X, what specific features mattered, what drove decisions.

What Gong showed: That Competitor X's name came up a lot.

The depth wasn't there.

Month 3: Feature Request Mirage

Gong flagged "Feature Y requested 23 times."

I listened to those 23 mentions:

• 8 were prospects asking "Can you do [Feature Y]?" AE said "Yes" and moved on (no discussion of why it mattered)
• 6 were prospects asking about it because competitor had it (checking boxes, not indicating real need)
• 5 were AEs proactively mentioning we had it (not customer-initiated)
• 4 were actual requests with context on why they needed it

Only 4 of 23 "feature requests" were genuine needs with context.

Gong counted mentions. Manual interviews would have uncovered actual importance.

Month 4-6: The Pattern Recognition Problem

Gong excelled at finding what I was looking for.

I set up trackers for specific keywords:

• "Integration"
• "Implementation time"
• "Ease of use"
• "Competitor X"

Gong found every mention of these keywords.

The problem: Gong only found what I searched for. It couldn't surface patterns I wasn't expecting.

Example:

In manual interviews, I discovered customers cared deeply about "team collaboration during launches."

I hadn't set up a Gong tracker for this because I didn't know to look for it.

Gong couldn't tell me: "Hey, customers keep mentioning this thing you're not tracking."

Manual interviews surfaced unexpected insights. Gong confirmed expected patterns.

Platform 2: UserTesting ($15K)

After realizing Gong was surface-level, I tried UserTesting for more structured research.

The promise: "Run customer research studies in 48 hours. Get feedback from 50+ users faster than you can schedule 5 interviews."

Month 1: The Speed Test

I ran my first UserTesting study:

• Question: "What's the biggest challenge in your current product launch process?"
• Participants: 20 product marketers
• Time to results: 48 hours
• Cost: Included in subscription

Results came back fast:

• 12 participants mentioned "coordination across teams"
• 8 mentioned "tracking launch tasks"
• 6 mentioned "measuring launch success"
• 4 mentioned other challenges

This was fast. But the responses were surface-level:

"Coordination across teams is hard" (no detail on why or what specifically breaks down)

"Tracking launch tasks" (no context on what tasks, what tools they use now, what's missing)

Month 2: The Follow-Up Problem

The UserTesting responses raised more questions than they answered.

Response: "We struggle with launch coordination."

Follow-up questions I couldn't ask:

• What specifically breaks down? Is it communication? Timing? Ownership?
• What have you tried? What didn't work?
• How does this impact your launches?
• What would solve this for you?

With manual interviews, I could ask follow-ups. With UserTesting, I got one-time responses with no ability to dig deeper.

I ran follow-up studies to clarify, but:

1. Different participants each time (no continuity)
2. Still surface-level responses
3. Cost adding up ($750 per study)

Month 3: The Participant Quality Issue

UserTesting's participant pool was hit-or-miss.

Study: "How do you handle competitive intelligence?"

Participants included:

• 8 actual PMMs at B2B SaaS companies (target audience)
• 6 marketers with "product marketing" in title but doing demand gen (not relevant)
• 4 people who didn't seem to understand the question
• 2 who gave one-sentence answers

Only 8 of 20 participants provided useful insights.

With manual interviews, I recruited specific people. With UserTesting, I got whoever signed up.

The Manual Interview Alternative

After 6 months with Gong and UserTesting producing surface-level insights, I went back to manual interviews.

My approach:

• 2 customer interviews per week (8 per month)
• 45-60 minutes each
• Semi-structured (prepared questions but followed interesting threads)
• Recorded and transcribed (used free Otter.ai)

Time investment: 12 hours per month (vs. 5 hours reviewing Gong/UserTesting insights)

Insights quality: 10x better

Example: Understanding competitive losses

Gong approach:

• Gong flagged "lost to Competitor X" in 12 calls
• Listened to calls, heard surface reasons: "They chose Competitor X"
• No depth on why

UserTesting approach:

• Survey: "Have you evaluated Competitor X? What did you like?"
• Responses: "Good features," "Easy to use," "Better price"
• Surface-level, no specific insights

Manual interview approach:

• Asked: "Walk me through your evaluation of us vs. Competitor X."
• Customer explained detailed comparison: Specific features, implementation concerns, team preferences, political dynamics
• Asked: "What ultimately made you choose them?"
• Uncovered: They didn't choose based on features—they chose based on implementation speed (we quoted 6 weeks, Competitor X quoted 2 weeks)

This insight changed our sales approach. We improved implementation speed and started leading with it.

Gong and UserTesting would never have surfaced this. They showed what customers said, not why they decided.

When Automated Platforms Help

After a year with both platforms, I learned where they added value:

Gong is valuable for:

1. Competitive mention tracking (which competitors come up, frequency)
2. Topic trending (what themes are increasing over time)
3. Sales coaching (how AEs handle objections)
4. Conversation patterns (what questions prospects ask)

Gong is not valuable for:

• Understanding why customers buy
• Uncovering deep motivations
• Finding unexpected insights
• Following up on interesting threads

UserTesting is valuable for:

1. Quick validation (test messaging with 20 people in 48 hours)
2. First impressions (how do people react to new positioning?)
3. Usability testing (can people find what they need on website?)
4. Broad patterns (what do 50 people think about X?)

UserTesting is not valuable for:

• Deep insight generation
• Understanding complex buying decisions
• Following interesting threads
• Building relationships with customers

The Integrated Alternative

After proving automated platforms didn't replace manual interviews, I looked for tools that made manual interviews more efficient instead of trying to replace them.

What I needed:

• Easy scheduling (reduce friction to book interviews)
• Good recording and transcription (so I could focus on conversation, not notes)
• Simple synthesis (organize insights without complex platforms)
• Lightweight participant tracking (remember who I've talked to, when to follow up)

I dropped Gong ($30K) and UserTesting ($15K) and used:

• Calendly for scheduling ($0—had it already)
• Zoom for calls ($0—had it already)
• Otter.ai for transcription ($0 for basic tier)
• Notion for synthesis ($0—had it already)

Total cost: $0

Time per interview: Same (60 min)

Insights quality: Same as manual (10x better than automated platforms)

Another approach worth exploring: integrated PMM platforms like Segment8 that combine lightweight research tools with win/loss tracking and customer feedback in one place, solving the integration problem.

Example of integrated workflow:

1. Schedule win/loss interview via platform
2. Conduct interview (platform records and transcribes)
3. Tag key insights during interview
4. Platform automatically connects insights to:
   • Competitive intelligence (competitor mentioned → updates competitor profile)
   • Feature requests (feature mentioned → surfaces to Product)
   • Messaging validation (what resonated → informs positioning)

This solves the integration problem: Insights from interviews automatically feed into competitive intelligence, product feedback, and messaging—without manual synthesis across tools.

Reported cost advantage: Part of broader PMM platform (vs. $45K for Gong + UserTesting)

The Real Costs Comparison

Option 1: Automated Platforms (Gong + UserTesting)

• Licenses: $45,000/year
• Time reviewing automated insights: 5 hrs/week × 50 × $80/hr = $20,000/year
• Insights quality: Surface-level, broad patterns
• Total cost: $65,000/year
• Actionable insights: ~5-10 per year

Option 2: Manual Interviews Only

• Licenses: $0 (Zoom + Otter.ai + Notion)
• Time conducting and synthesizing: 12 hrs/month × 12 × $80/hr = $11,520/year
• Insights quality: Deep, actionable
• Total cost: $11,520/year
• Actionable insights: 40-50 per year

Option 3: Integrated Platform with Research Tools

• License: ~$6,000/year (as part of PMM platform)
• Time conducting research: 10 hrs/month × 12 × $80/hr = $9,600/year
• Insights quality: Deep, automatically connected to workflow
• Total cost: $15,600/year
• Actionable insights: 50+ per year with better integration

Automated platforms cost 4x more and generated 5x fewer actionable insights than manual interviews.

What I'd Tell My Past Self

If I could go back to when Product suggested buying Gong and UserTesting, here's what I'd say:

"Automated research platforms scale breadth, not depth. PMM needs depth."

Gong can analyze 1,000 sales calls. But it gives you surface mentions, not deep understanding.

Manual interviews with 20 customers give you actual decision-making context, motivations, and unexpected insights.

For product marketing, 20 deep interviews beat 1,000 automated call analyses.

The questions PMM needs to answer:

• Why do customers choose us vs. competitors? (requires deep conversation)
• What outcomes do they actually achieve? (requires context and follow-up)
• What messaging resonates? (requires testing and iteration)

Automated platforms answer: What do customers mention? (mentions ≠ importance)

The Uncomfortable Truth

After spending $45K on customer research automation, here's what I learned:

Most PMMs don't need expensive research automation platforms. They need consistent manual research integrated with their actual workflow.

The problem isn't scaling research quantity. It's:

• Getting deep insights that drive decisions
• Following interesting threads
• Understanding context and motivation
• Connecting insights to actions

Automated platforms optimize for breadth. What PMM needs is depth.

That's why I cancelled Gong and UserTesting after one year. Not because they weren't powerful—they generated tons of data. But data ≠ insights.

The most valuable insights that changed our positioning, messaging, and sales approach all came from manual interviews where I could ask follow-up questions and understand context.

Save the $45K. Invest in consistent manual research with lightweight tools that make synthesis easy, not automation that generates surface-level data.