Dmitry Sergeev
Upwork proposal · Consumer market research & customer discovery

Who adopts first, and why they come back

Dmitry Sergeev · customer discovery and GTM for consumer and two-sided products

Your brief asks for a launch decision, not a market report. Which people have a strong enough reason to try a new communication product? What makes the other side show up? And how do you reach both?

I've co-founded 10 companies and run a few hundred interviews for my products and clients. At RichAds, a two-sided ad network, interviews and funnel data told us to stop chasing "advertisers" and go after affiliates and brands in iGaming, Nutra and Crypto. We went from $0 to $1M MRR in 16 months.

  • $0 to $1M MRR on a two-sided network
  • A few hundred customer interviews
  • 1,000+ early adopters before launch
  • Segment scorecard for a dating app launch

Your 5 questions, answered on this page

Remote from Batumi, Georgia (GMT+4) · English C1, Russian native

Your brief

How I read it

You already have a product. The open question is where it gets used for real, so I'd start from three working rules.

01

It takes two people to use it

A communication product is empty for the first user until someone else joins. For every segment I check whether the first person gets anything before the other side arrives, who invites whom, what the first good interaction looks like, and why they open it again next week.

02

The workaround is the proof

People who really have the problem have already patched it: a WhatsApp group, a Discord server, three apps taped together. The patch shows the need is real, and it shows what you have to beat.

03

Every score shows its evidence

Next to each score you'll see the quotes and numbers behind it, and how sure I am. Interviews show patterns and reasons. They don't size a market, so each segment also gets a survey and a small fake-door ad test with a real CTR and cost per signup.

Assumptions I'll test first

  1. Who the two sides really are, and which one is harder to get.
  2. Whether one person gets value before the other side joins.
  3. Whether today's workaround is good enough for most people.
  4. Who pays, for what, and how often.
  5. What people say they'd do, against what they do when you send them a link.

The finding you get

Weak

"Go after students. They use messengers a lot."

Strong

"Group X hits problem Y every week and patches it with workaround Z, which costs them time or money. We talked to both sides, we know where to recruit them, and here is the test that shows whether they switch."

Format illustration, not a guess about your product.

1 · Two research projects

Two projects where research picked the audience

In both, the useful part of the decision was saying no to a segment that looked big.

RichAds logo

RichAds

My company · two-sided ad network · 2018–2020
Context
An ad network: advertisers buying traffic on one side, publishers selling it on the other.
Open question
Which advertisers do we win first? "Advertisers" is everyone, and a message for everyone lands with nobody.
What I did
Customer interviews plus funnel data. I turned the interviews into a 9-stage journey map with buyer quotes, barriers and KPIs for every stage.
What the evidence showed
The scalable segments were affiliates and brands in iGaming, Nutra and Crypto. Their job: "my Google and Facebook costs keep climbing, I need a source that converts at a cost I can defend." The map also showed where the money leaked: between sign-up and first deposit.
Decision
Repositioned around profit protection and went after those verticals. Dropped: advertisers as one broad segment.
Result
$0 to $1M MRR in 16 months. 20,000+ clients. CAC down 35%.

Between

Client · staged-reveal dating app · NYC and Austin · 2026
Context
Between replaces the swipe with guided sessions. Two people discover each other step by step, and a conversation opens only if both choose to continue.
Open question
Who to win first in a crowded category, and where, so that the first users don't land in an empty app.
What I did
Researched the US matchmaking market: paid matchmakers, vetted apps, cultural networks, intent-based apps. Ran discovery interviews with daters. Took apart five predecessors to see which launch mistakes were already documented.
What the evidence showed
Online audiences were cheap to reach but had no density, so a new user would open an empty feed. Closed groups, where both sides meet every week, had both reach and density.
Decision
Scored 6 segments on demand, frequency, monetization, access and density. First segment: women 25 to 34 who are tired of the apps, inside closed groups in Williamsburg and Greenpoint. Dropped for now: online communities, college students and creator audiences.
Launch test
Hand-picked 250 local nano and micro creators and shipped 2 test collabs, tracked to signups through tagged links.
Where it stands
The scores are marked pre-validation on purpose. An 8-week plan with 4 gates replaces each one with measured data.
Ultra Boxing VR logo

Bonus, consumer: Ultra Boxing VR

I grew 1,000+ early adopters in Discord before launch. We tested stealth builds with them, and their feedback shaped the roadmap. Paid acquisition dependency went down 30%.

2 · Finding early adopters

I look for the workaround

Early adopters have the problem, know they have it, and already hacked something together to deal with it.

Step 1Has the problem
Step 2Knows it
Step 3Already hacked a workaround

A WhatsApp group, a Discord, three apps taped together. The workaround is the proof.

Alternatives map

Competitors plus the workarounds people actually use, mined from reviews, Reddit, Discord and Telegram.

JTBD switch interviews

What they did the last time the problem hit, and what it cost them.

Value proposition canvas

One per segment, for both sides of the platform.

Journey map with real quotes

Including the moment the other side has to show up.

Atomic network test

From The Cold Start Problem: which side is hard, and the smallest group that works on its own.

3 · Recruiting and interviews

Yes, I recruit and interview on my own

I've run a few hundred interviews for my products and clients. I find the people, screen them, book them and run the calls myself.

CommunitiesLocal creatorsScreener ads on Meta and TikTokPanels like RespondentClay, if one side is B2B
  1. Find where they gatherCommunities, creators and panels where both sides already are.
  2. Screen for recent experienceThe problem happened to them lately, so they remember what they did.
  3. Invite one by oneA personal message, not a mass blast.
  4. Book and get consentScheduling, reminders and consent to record.
  5. Over-recruitExtra people booked to cover no-shows.
  6. Incentives at costPaid to participants separately from my fee.

I confirm the channel mix once we agree on geography and screening criteria.

4 · Methodology, timeline, budget

4 weeks, 4 milestones, $6,000 fixed

Each milestone ends with a short readout. You can stop or change direction at any of them.

  1. Week 1

    Map the assumptions

    Product walkthrough and a call with you. Assumption map, alternatives map, a long list of 10 to 15 segments, screeners.M1 · $2,000 20–30 h
  2. Week 2

    Talk to people

    Recruiting and the first half of 20 to 25 depth interviews across both sides. A 150-person survey and fake-door ads go live, $50 to $100 per segment.M2 · $1,300 12–20 h
  3. Week 3

    Score the segments

    The rest of the interviews. Survey and ad results, so demand comes with a CTR and a cost per signup. Segment scorecard, personas and use cases.M3 · $1,300 13–20 h
  4. Week 4

    Pick and plan

    Top 2 to 3 segments, an acquisition plan for each, and an 8-week GTM test roadmap with kill rules. Live readout.M4 · $1,400 15–20 h

A behavior test inside your build

The product already exists, so I'd rather watch people use it than ask them about it. If the test build allows, interviewees get an invite, and I track six steps. We pick the exact actions in week 1.

Onboarding done→Key action→Invite sent to the other side→Invite accepted→First useful interaction→Came back

The segment scorecard

Your four criteria plus density. Each score sits next to its quotes and numbers, with a confidence level: high, medium or low.

CriterionWhat I look atEvidence behind the score
DemandA recent, painful trigger, and an active search for something betterInterviews, survey, fake-door CTR and cost per signup
FrequencyHow often the situation comes backThe last 3 times, in the interviewee's own words
MonetizationWho pays and for what, in three layers: what they already pay for workarounds, what they say, what they do when they see a priceCurrent spend, stated willingness to pay, fake-door with a price
AccessibilityCost and speed to reach 100 of themRecruiting cost and response rate from weeks 2 and 3
DensityWhether both sides already gather in one place we can enterCommunity map and the atomic network test

Budget and hours

PhaseWeeksHoursMilestone
Product audit, hypotheses, research plan18–12 hM1 · $2,000, with the row below
Desk research, alternatives, segment map112–18 hM1
Recruiting, interviews and processing2–325–40 hM2 · $1,300
M3 · $1,300
Synthesis, prioritization, GTM roadmap3–415–20 hM4 · $1,400
Total460–90 h$6,000 fixed
At cost, on topParticipant incentives, about $1,000 to $1,500. Ad tests, $500 to $1,000.
Paid through UpworkFour milestones, released as each one is delivered.
What the 4 weeks assumeProduct access in the first days and people we can recruit online. If recruiting looks harder, I'll tell you in week 1, before the interview money is spent.

5 · Deliverables

What you get, and what it looks like

Your six deliverables, mapped to what I hand over. Below them, two documents I've already made.

Your deliverableWhat you get
1. Market segmentation and opportunity analysisA long list of 10 to 15 segments, the alternatives map, and the segment scorecard on 5 criteria
2. Findings backed by real data and user insightsNotes from 20 to 25 interviews with quotes, the 150-person survey, fake-door CTR and cost per signup per segment, and the behavior test funnel
3. Prioritized audiences with reasoningTop 2 to 3 segments, each with the evidence for and against it, plus the segments I'd drop and why
4. Personas and high-potential use casesPersonas for both sides, built from the interviews, with the trigger moment for each use case
5. How to reach and acquire themAn acquisition plan per segment: where they gather, the first channel, the message, and the cost per signup seen in the ad tests
6. A practical GTM testing roadmapAn 8-week roadmap: one test per week, a success metric for each, and kill rules
Between segment scorecard: six segments scored on demand, frequency, monetization, access and density

Between · 2026

Segment opportunity scorecard

Six entry segments scored on five criteria, a teardown of five predecessors, five launch rules and an 8-week plan with four gates. It picked one first segment and set three aside.

RichAds customer journey map: nine stages with steps, thoughts and barriers

RichAds · from customer interviews

Customer journey map, 9 stages

From "where is cheaper traffic?" to loyal spender. Each stage has buyer quotes, barriers, KPIs and a CRM status, and the map shows where the money leaked.

Experience

Consumer and two-sided products

Upwork logo

2023 – present

Freelance on Upwork

GTM and growth lead

Fractional growth for startups, apps and marketplaces.

  • Led growth for two-sided marketplaces in dating, jobs and sports. Separate funnels and messages for each side, optimized for completed matches, applications and bookings instead of signups.
  • Customer discovery and segment research for Between, a staged-reveal dating app in NYC and Austin.
  • Ran 150 creative tests in 3 weeks on Meta. Found 10 winners and cut CPL by 38%.
Telegang Agency logo

2024 – 2025

Telegang Agency

Head of Growth & Strategy

Telegram growth agency for media outlets and brands.

  • Acquired 2M+ targeted subscribers across the client portfolio through paid and organic funnels.
  • Kept retention above 50% with audience segmentation and lifecycle automation.
  • Built subscriber economics models that tied acquisition cost to ad revenue per subscriber.
Ultra Boxing VR logo

2022 – 2023

Ultra Boxing VR

Co-founder, Producer & Growth Lead

VR boxing and fitness game, now live on Meta Quest.

  • Grew a community of 1,000+ early adopters in Discord before launch and tested stealth builds with them.
  • Fed their feedback into messaging tests and the product roadmap. Paid acquisition dependency down 30%.
  • Rewrote the landing page around what a player feels in the ring. Conversion up 25%.
RichAds logo

2018 – 2020

RichAds

Co-founder & Director of GTM & Revenue

Push, popunder and native ad network: a two-sided marketplace of advertisers and publishers.

  • Took the company from $0 to $1M MRR in 16 months by growing both sides of the marketplace.
  • Interviews and funnel data picked the segments: affiliates and brands in iGaming, Nutra and Crypto. I built a 9-stage journey map from those interviews.
  • Repositioned the brand around profit protection. Cut CAC by 35% through funnel testing, lead scoring and better MQL-to-SQL routing.
  • Onboarded 20,000+ clients. Built a partner network that drove 40% of revenue.
Platform.io logo

2013 – 2018

Platform.io

Head of International Business Development

White-label adtech SaaS.

  • Led market entry into North America and APAC with regional GTM playbooks.
  • Turned technical specs into value propositions that non-technical enterprise buyers could act on.

10 companies co-founded in total, $100M+ in combined revenue.

Methods

Assumption mappingAlternatives mappingJTBD switch interviewsValue proposition canvasJourney mappingScreener surveysFake-door testsSegment scoringAtomic network test

Tools

RespondentClayMeta AdsTikTok AdsGA4HubSpotClaude

Languages

English C1. Russian native.

Before we start

Questions for our first call

The answers change the scope, so I'd rather ask them before the first milestone than after.

  1. What does the product do, and who are the two sides of an interaction?
  2. Who has tested it so far, and what did they actually do in it?
  3. Which product and audience decisions are you ready to change after the research?
  4. What are the launch geography, languages and planned monetization?
  5. Will I get access to the product and analytics, and can I invite people into the test build?
And the one that matters mostIf the data shows your first audience doesn't need the product in its current form, are you ready to change direction?

Let's find your first users

Reply here on Upwork and we'll set up a call. Bring your current hypotheses, and I'll tell you which ones I'd test first.