Acquisition◆Retention◆Expansion◆Partnership◆HubSpot Systems◆ICP Architecture◆Pipeline Infrastructure◆CRM Infrastructure◆Revenue Operations◆B2B◆ Acquisition◆Retention◆Expansion◆Partnership◆HubSpot Systems◆ICP Architecture◆Pipeline Infrastructure◆CRM Infrastructure◆Revenue Operations◆B2B◆
Revenue Operations · Funded Startups · Enterprise · Service Providers · 10+ Countries

I build it,
run it, and keep it working.

Most revenue problems are not strategy problems. The strategy is usually fine. What's broken is the infrastructure between the strategy and the result — the CRM nobody trusts, the handoff that drops context at close, the ICP that stopped being accurate two years ago and nobody updated. Funded startups, enterprise sales teams, service providers scaling past referrals — the shape of the business is different. The failure points are almost always the same. I build go-to-market systems for acquisition, and the RevOps infrastructure that makes retention and expansion reliable — then keep all of it running, so your team is closing deals, not chasing data, rebuilding context, or doing work that should already be automated.

$48M+
Pipeline Generated
50+
B2B Companies
12+
Countries
+30%
Avg. Retention Lift
Arushi Mittal
How Revenue Gets Built
Revenue isn't one problem.
It's three — in sequence.

Most companies have invested unevenly across these three motions — and the imbalance compounds quietly. A pipeline that looks healthy until it doesn't. Retention that holds until a handoff breaks it. Expansion that never gets systematised because new logo always feels more urgent.

"The gap is almost never strategy. It's the infrastructure between the strategy and the result."

A CRM that records activity without surfacing intelligence. A handoff that loses context at the exact moment it matters most. An ICP built in year one that's been suppressing pipeline ever since. These aren't exceptional failures — they're the default state for most revenue teams.

I work across all three motions — building the infrastructure that makes each one visible, reliable, and not dependent on any single person to hold it together. The details are on the right.

The revenue lifecycle
01
Acquisition
Go-to-market strategy, ICP architecture, product-market fit validation, and lead scoring — built to get in front of the right companies at the right time.
02
Retention
Keeping what was won — starting from the handoff, not the renewal conversation.
03
Expansion
Growing accounts deliberately — not by waiting for them to ask for more.
Results in Practice
Harpin AI ↗
"Testimonial placeholder — replace with a real quote from the client about the engagement, the results, or what changed."
Client Name · Role
Stone Cold Leads ↗
"Testimonial placeholder — replace with a real quote from the client about the engagement, the results, or what changed."
Client Name · Role
Who I Work With
Chief Revenue Officer

"I can't get a straight answer on pipeline velocity. And I shouldn't have to ask."

Forecasting is still a manual exercise. Closing velocity, stage duration, at-risk signals — none of it surfaces without someone pulling it together. You're building the picture yourself when the infrastructure should already be doing it for you.

Read more →
Sales Leader

"My best reps are spending half their week on work that isn't selling."

Manual research, record updates, list building — tasks that eat the hours your team should be spending qualifying accounts and having real conversations. The capacity exists. It's buried under process that was never systematised.

Read more →
CEO / Founder

"I hired a sales team. Revenue isn't scaling the way I expected."

What you built in the early days was a revenue motion that ran through you — your judgment on which accounts to prioritise, your instinct on when to push, your read on a deal. That motion doesn't transfer automatically. It needs to be codified into systems your team can operate without you in every conversation.

Read more →
Chief Revenue Officer

You shouldn't need to chase your own data to understand what your revenue team is actually doing.

7AM. Board deck due at noon.

"It's Monday morning. The board deck is due by noon."

You open the CRM at 7am. The pipeline number is there — but you already know it's wrong. Two deals that haven't moved in six weeks are still sitting at the same stage because nobody updated them. A rep who's been discounting to close looks strong on paper. The segment you've been worried about is producing volume, but the churn it's generating won't show up until next quarter. You do the mental adjustment — the one you always do — and you arrive at the number you'll actually present. Not the one in the system.


Downstairs, your CFO is asking investors why pipeline conversion has been inconsistent for three quarters running. Your CEO wants a clean story. The board wants confidence. And you're the one standing between all of them and a dataset you don't fully trust — with ninety minutes to build a narrative around it.


This is not a data problem. Every piece of information you need exists somewhere in your CRM. It's an architecture problem. The system was never built to surface intelligence automatically — it was built to track activity. And nobody has drawn the line between what the team does and what you can actually see.

Q2? Q3? ? "What does Q3 look like?"

"Your investors are asking what next quarter looks like. You don't have a clean answer."

The question every investor asks is simple: how much revenue are you going to make, and how do you know? The honest answer — for most revenue teams — is that the forecast is a blend of gut feel, historical averages, and a pipeline number that's been manually adjusted by whoever built the slide. It looks like a model. It doesn't behave like one.


Without reliable stage velocity data, you can't tell your investors with any confidence what closes in the next 90 days. Without visibility into deal health, you can't identify which accounts are at risk before they become churned logos. Without a structured view of what's working in acquisition and what isn't, you can't explain why growth is compounding or why it's slowing — you can only describe the outcome after it's already happened.


Investors aren't asking for certainty. They're asking for a system that produces reliable signals — one where you can say: here is what the data shows, here is what we believe it means, and here is the early warning infrastructure that will tell us if we're wrong before it's too late to act. That system doesn't exist by default. It has to be built.

guesswork no signal infrastructure Build the system. See the path.

"The reporting I need should exist without me requesting it."

A CRO should not be configuring dashboards at 7am. Should not be chasing their own data to understand what the team is doing. Should not be walking into a pipeline review carrying a number they've already adjusted in their head because the system can't be trusted at face value.


When your team knows that data quality isn't enforced, they optimise for optics rather than accuracy. Stage updates get made to satisfy review cycles, not to reflect deal reality. The pipeline inflates in ways that are difficult to challenge without creating friction. And you end up carrying the honest number privately while presenting a managed version publicly — an enormous cognitive and political burden that should not be part of this job.


What's missing isn't effort. It's the infrastructure underneath — the stage definitions that mean something, the workflow logic that enforces data quality without requiring manual discipline, the automated surfaces that flag anomalies before you have to go looking for them. That's the work that makes Monday mornings different.

Product CRO CSM Sales Arushi

"I am the connective tissue between your product, your CS team, your sales team — and what you can actually see as a CRO."

Revenue visibility doesn't fail because people aren't working. It fails because the four teams generating the signals — Product, CS, Sales, and the CRO's own reporting layer — were never systematically connected. Product ships something. Sales doesn't know how to position it. CS inherits accounts without context. The CRO sees numbers that reflect none of what actually happened.


My role is to sit at the centre of those four functions and build the infrastructure that makes them legible to each other. What Product ships gets reflected in Sales messaging. What Sales closes gets transferred cleanly to CS. What CS observes about health and churn feeds back into the CRO's pipeline view. The information flows in both directions — and the CRO stops being the person who has to manually reconstruct the picture every time they need to make a decision.

"The CRO who has to chase their own numbers has an infrastructure problem, not a reporting problem. The fix is upstream — in the CRM logic, the stage definitions, the handoff architecture, and the automated surfaces that should be delivering intelligence without anyone being asked to go find it."

The diagnostic conversation starts with the question you've already been asking internally: what would it take to actually trust the number? Thirty minutes. No deck required.

Sales Leader

Your team's capacity isn't the constraint. What they're spending it on is.

THU 14 Q3 ? ↓ close % ? Pipeline review. Numbers don't add up.

"It's Thursday afternoon. The pipeline review starts in an hour."

You pull the CRM. Three reps have deals sitting at the same stage they were in two weeks ago — nobody updated them. Your best rep has seven opportunities open but you have no visibility into which ones are actually progressing. The forecast you're about to present is a number you've assembled from memory, intuition, and a spreadsheet you built yourself because nothing the system produces is clean enough to share directly.


You already know what the conversation will look like. You'll ask each rep for a deal-by-deal update. They'll tell you what they remember. You'll write it down. You'll leave with a slightly more informed version of the picture you had going in — and next week you'll do it again, because there's no system in between that's doing the tracking work for you.


Your team is not underperforming. Your infrastructure is. The difference matters because only one of those is fixable this quarter.

capacity ceiling CRM log research lists automated your rep now . your rep after Remove the ceiling. Free the rep.

"The manual overhead is the ceiling on how many accounts they can meaningfully run."

A rep spending two hours a day on administrative work — logging calls, building lists, researching accounts from scratch, writing follow-up notes in a format nobody standardised — is running fewer accounts, with shallower engagement, than a rep whose administrative layer has been systematised. This is arithmetic, not opinion.


What makes it worse: the reps who perform best in these environments aren't necessarily the strongest sellers. They're the ones who've quietly built their own workarounds. Their own shortcut for logging. Their own research template. Their own follow-up system. Which means your top performers are carrying institutional process failures the rest of the team is drowning in — and when those reps leave, they take those workarounds with them.


That's not a talent problem. It's an infrastructure problem wearing a talent problem's clothing.

admin selling → CRM updates → list building → research → note-taking → formatting ~60% of day → qualifying → stakeholder map → timing reads → objections → closing ~40% ⇄ flip the ratio Free judgment. Systematise the rest.

"I want them focused on judgment. Not data entry."

The work that separates good salespeople from average ones is qualitative. Reading a stakeholder map correctly. Knowing when to push for a timeline and when to give a deal room to breathe. Identifying which objections are real and which are deflection. Knowing which contact is the genuine internal champion and which one is enthusiastic but carries no authority. None of that can be automated. All of it requires time, attention, and mental space.


That mental space gets crowded out when a rep's day is structured around quantitative throughput — calls logged, emails sent, records updated. When I work with a sales team, the goal isn't to make them faster at administrative work. It's to remove as much of it as possible so the hours it was occupying get redirected to the qualitative work that actually determines whether accounts close.


Your reps don't need to become better sellers. They need to be given the conditions where their selling ability is actually the constraint — rather than spending it competing with a to-do list that should have been systematised before they walked in the door.

cold 3 weeks silent Deal went cold. Nobody flagged it. You find out after. Not during.

"When something goes wrong in the pipeline, I'm the last to know."

An account goes quiet for three weeks. No flag. A rep's close rate has been dropping for two months — it shows up in the quarterly review, not in week six when there was still time to intervene. A deal that's been sitting at the same stage for a month isn't at-risk in anyone's system — it's just there, static, until someone happens to notice.


You find out about problems when they've already become losses. And you're expected to have coaching conversations based on your own memory and observation — because there's no system surfacing the right information to you automatically, at the right moment, in a form you can act on.


That's not a management failure. It's an infrastructure gap that makes good management much harder than it should be. The signals exist. They're just not being routed to you.

"The best sales leaders I've worked with aren't asking their team to do more. They're asking their systems to do more — so that the team's attention goes where human judgment genuinely matters and the infrastructure handles everything it doesn't need a person to do."

The first conversation is a clear-eyed look at where your team's hours are actually going — and what it would take to redirect them to the work that moves the number.

CEO / Founder

You raised the funding. You're still doing everything yourself.

BD OPS DELIVERY CX SALES Series A closed ✓ One person. Five jobs.

"You raised the round. The job description didn't change."

The deck said you were going to hire a BD team, a sales function, an ops lead. The money is in the bank. But it's six months later and you're still the one writing the outreach. Still on every first call. Still managing delivery escalations when a client is unhappy. Still doing the CX work because nobody else knows the accounts well enough yet. Still figuring out the operational infrastructure that was supposed to exist before you started scaling into it.


You are simultaneously the BD team, the sales team, the ops function, the client success team, and the CEO. None of those roles are getting the attention they deserve because you're splitting yourself across all of them — and the company is growing at the speed of your personal bandwidth, not at the speed it should be capable of given the capital you've just raised.


The problem isn't that you're bad at delegating. It's that there's nothing built to delegate into. The systems, the processes, the handoffs — they don't exist yet. So every time you try to hand something off, it bounces back.

deal ? ? You are the pipeline.

"You are the pipeline. And that's the problem."

Every deal in your pipeline either came through you, got saved by you, or is waiting on you to move it forward. That's not a BD function — it's a founder doing BD between everything else they're supposed to be doing. It means your pipeline grows when you have time for it and stalls when you don't. Which is almost always.


The deeper problem: the knowledge required to run BD effectively — which companies to target, how to frame the value proposition, what triggers a decision at the buyer level — lives entirely in your head. It was never codified into a repeatable system that someone else could execute. So even when you do hire a BD or sales person, they spend the first three months trying to extract from you what should already exist in a document, a CRM, a target list, a message framework.


The pipeline should not depend on your personal availability. The targeting logic, the messaging, the contact strategy — these need to exist as infrastructure that runs without you in the room.

? ? team hired. still waiting on you. Leverage that loops back to you.

"I hired people to take this off my plate. It's still on my plate."

You hired a salesperson. They're three months in and still asking you which accounts to go after. You hired an ops person. They built a process but it only works when you're available to answer questions about the edge cases. You hired a CS person. Every escalation still finds its way back to you because they don't have the institutional context to resolve anything beyond the surface level.


This is the codification problem. What you know — which accounts matter, how to read a deal, what good client onboarding looks like, how to handle a difficult renewal conversation — was never written into a system. It lives in your decisions and instincts, accumulated over hundreds of conversations. And it doesn't transfer to a new hire through an onboarding deck and a few shadowed calls.


The result: a team that's technically in place but functionally dependent on you for anything that requires judgment. You've added headcount without adding capacity. And you're more stretched than you were before the hires, because now you're managing people on top of still doing the work.

before after system founder: CEO Build the system. Step back from it.

"I know what good looks like. I just need it built."

Most founders at this stage aren't confused about what the operation should look like. They've been in enough conversations, made enough hires, and felt enough friction to know exactly what's missing. What they don't have is the time or the specific infrastructure expertise to build it — and the company keeps moving forward on top of gaps that are getting more expensive the longer they remain unaddressed.


The right sequence: build the system before hiring people to run it. Define the ICP logic, the CRM structure, the handoff processes, the onboarding framework, the reporting surfaces — and then hire into a structure that already exists. Not the other way around. A hire walking into undefined infrastructure spends three months figuring out what they're supposed to build, from a brief that doesn't exist, inside a company that isn't sure what it needs.


My role is to build the architecture that lets you step back from the execution — so you're running the company, not substituting for the systems it should already have.

"The founders who scale successfully are almost always the ones who at some point made the decision to treat the revenue motion as an operational discipline rather than a personal skill. That transition — from the founder doing it to a system doing it — is not automatic. It requires explicit work. And it almost never happens until the cost of not doing it becomes undeniable."

The first conversation is usually the most clarifying thirty minutes a founder has had about their revenue operation in months. No preparation required — just an honest account of where it is and where it isn't working.

Acquisition
01 · Acquisition
The precision problem.
Not the volume problem.

Most pipeline shortfalls aren't about doing more. They're about reaching the right companies, at the right time, with a message that reflects their actual situation — not a template built around a job title. I build the go-to-market strategy and acquisition infrastructure that changes that ratio.

GTM Strategy ICP Architecture Lead Scoring Market Scoping
"The ICP you built in year one is almost certainly suppressing your pipeline in year three. The data to fix it exists inside your own CRM. Most companies never look at it."

Most targeting decisions are made once — during the scramble to find early customers — and never revisited. Three years later, the company is running sequences to the same titles, same company sizes, same verticals. Everything looks right on paper.

The data inside the CRM tells a different story. The accounts that close fastest, discount least, and renew most reliably share specific characteristics that aren't reflected in the current targeting criteria. The work starts there.

The Problem
ICP DRIFT OVER TIME YEAR 1 ICP SMB · 50-200 emp YEAR 3 REALITY Mid-market · 200+ drift gap Pipeline looks active. Converts poorly. CRM DATA closed-won patterns untouched the fix starts here

"The ICP was built in year one and nobody has touched it since."

Most companies build their ICP during the early scramble for customers — targeting whoever would talk to them, refining it based on whoever closed, and then filing it away as a done item. Three years later the company looks completely different. The product has evolved. The customer profile has shifted. The accounts that close fastest today bear little resemblance to the early adopters who validated the idea.


But the targeting criteria hasn't changed. The sequences still go to the same job titles, the same company sizes, the same verticals. The team assumes the motion is working because it was working — and misses the compounding cost of reaching the wrong companies at scale. A pipeline that looks active but converts poorly is almost always a targeting problem, not a messaging problem. The data to fix it already exists inside the CRM.

VOLUME VS PRECISION HIGH VOLUME ✗ ✗ ✗ ✗ ~ 1% conversion ICP SCORED fit: 92 fit: 88 ✓ conv. ✓ conv. higher conversion rate

"Sending more isn't the answer. Sending to the right companies is."

The default response to a pipeline problem is almost always volume. More sequences. More contacts per account. More touchpoints per week. The logic feels sound: if a certain percentage of outreach converts, more outreach means more conversions. But the percentage only holds if you're reaching the right companies to begin with — and most teams are not.


A large proportion of outreach is going to companies that would never buy regardless of the message, the timing, or the number of follow-ups. The firmographic criteria are too broad, the trigger signals aren't being used, and the contact strategy isn't built around buyers who are actually in a position to make a decision. Increasing volume into a broken targeting model doesn't improve conversion — it just increases the scale at which you're reaching the wrong people.


The fix is precision: fewer companies, better qualified, reached at the right moment, with a message built around their specific context. That's what changes the ratio.

GTM VALIDATION FLOW SEGMENT DEFINITION who, why, when PMF VALIDATION test before building doesn't resonate iterate segment resonates → build GTM INFRA PIPELINE validated signal

"The GTM motion was built for the market you were in. Not the one you're entering."

Go-to-market strategy is usually treated as a one-time exercise — built during the early product phase, validated against the first cohort of customers, and then left largely intact as the company grows. The problem is that markets shift. The buyers who were accessible in year one aren't the same buyers who hold budget in year three. And the motion that worked in one geography rarely transfers to another without deliberate adaptation.


Before any outreach infrastructure is built, the motion needs to be validated. Which segments are genuinely reachable. What the buying trigger actually looks like at the ICP level. Whether the problem the product solves is the problem the target buyer feels acutely enough to act on. GTM strategy is the work that determines whether everything downstream is worth building — and most companies skip it in favour of just starting the sequences.

ACQUISITION SYSTEM firmographics intent signals tech stack ICP SCORING ENGINE closed-won analysis TRIGGER LAYER hiring · funding · expansion CONTEXTUAL MESSAGE specific · timed · relevant QUALIFIED CONVERSATION right company · right time

"What the system produces when it's built correctly."

When the targeting logic is built from closed-won data rather than assumption, the companies in the sequence are the ones that actually buy — not a broad approximation of them. When lead scoring is based on firmographic fit plus real-time trigger signals, outreach lands at the moment a company is most likely to be in a buying conversation rather than six months before or after it.


When the contact strategy is built around what the buyer actually experiences — their growth stage, their recent hiring activity, their technology stack, their competitive context — the message isn't a template. It's a reflection of their specific situation. And that is the difference between a sequence that gets deleted and one that starts a conversation worth having.


The goal is not more pipeline. It's a pipeline where the companies in it are the right ones — which means fewer deals that stall, shorter sales cycles, and accounts that stay.

What I Build
01

Go-To-Market Strategy

Before any outreach infrastructure is built, the motion needs to be right. Which segments, which entry point, which problem to lead with, in which order. GTM strategy is the prerequisite — not the afterthought.

02

Product-Market Fit Validation

For companies entering new markets: testing positioning and segment assumptions before committing to infrastructure. Know what's resonating before building systems around it.

03

ICP Architecture & Lead Scoring

Closed-won analysis to identify which accounts actually close, retain, and expand — translated into a scoring model the CRM can act on. Targeting criteria built from reality, not year-one assumptions.

04

Market Scoping & Entry Infrastructure

Database scoping for new geographies before any commitment is made. Identifying where the addressable market actually is and whether the current motion translates before investing in it.

05

Contact Strategy & Messaging Architecture

How to reach the right buyer, at what moment, with what framing — built around their actual context, not a job title and a template. The message earns the conversation; it doesn't just announce one.

06

Pipeline Visibility & Acquisition Reporting

Source attribution, conversion by ICP segment, stage velocity — built into the CRM so the data surfaces automatically. A pipeline view that tells you what's working, what isn't, and where the constraint is.

Ready to look at what your pipeline is actually telling you?
The diagnostic starts with the data you already have.
Retention
02 · Retention
Churn is a handoff failure. Not a CS failure.

In most churn cases, the failure happened before CS was ever involved. Context died at the handoff. The 60-day onboarding window was treated as administration rather than the most important commercial period in the entire lifecycle. I build the infrastructure that fixes that.

Handoff Systems Onboarding Architecture Health Scoring Renewal Visibility
"The decision to renew is largely made in the first 60 days — not in the weeks before the renewal date. By the time you're running a renewal conversation, the client has already decided."

Open the last ten churned accounts in your CRM at the point of close. In most cases: company name, deal value, close date, a notes field written at 11pm. The CS team inherited that record and was expected to pick up the relationship as if they were present for every conversation.

They weren't. That's not a CS failure. That's a handoff failure with a 12-month lag.

The Problem
THE HANDOFF GAP DEAL CLOSED sales team exits WHAT CS GETS company name deal value · close date ... that's it WHAT WAS LOST stakeholder map promises made concerns raised CS rebuilds context from scratch 60-day window ticking CHURN RISK ↑ fix: structured handoff Context lost at close = churn 12 months later.

"The churn was decided at close. Not at renewal."

When a deal closes, the sales team moves on. What they leave behind is almost never sufficient for CS to pick up the relationship at full context. The stakeholder map, the objections that were worked through, the promises made about what the product would do, the concerns that weren't fully resolved — none of it is in the deal record. It lived in conversations, email threads, and the sales rep's head.


CS inherits a shell. They're expected to build a relationship as if they were present for every sales conversation. They weren't. So the first 60 days become a period of re-establishing basic context rather than building value — and the client, who expected a seamless transition, starts to wonder whether the company is as competent post-sale as it seemed pre-sale.


That is when the renewal decision gets made. Not at the 11-month mark when someone sends a renewal proposal.

HEALTH SIGNAL FLOW login freq. support tickets NPS signal HEALTH SCORE ENGINE automated · real-time AT-RISK FLAG CS intervention HEALTHY renewal queue RETENTION ↑ acted on before it's too late

"The at-risk signal exists. It's just not being routed anywhere."

Login frequency drops. Support ticket volume increases. An NPS response comes back at a 6. A key stakeholder goes quiet for three weeks. All of these are observable signals — data the system already holds — but in most organisations they sit in separate tools, are reviewed inconsistently, and are never combined into a coherent picture of account health until someone decides to look.


By the time CS notices that something is wrong, the client has already mentally moved on. The conversation about retention becomes a conversation about saving an account that was lost months earlier — at the handoff, or in a first onboarding call that didn't build enough confidence, or in a QBR that nobody scheduled because no system enforced it.


The infrastructure I build routes those signals automatically. CS sees the health score without having to check five tools. At-risk flags surface before the silence becomes a cancellation notice.

RENEWAL AS A MOTION M0 M3 M6 M9 M12 renewal without system ↑ M11: first renewal touch with system onboard QBR health ck renewal Q ✓ 90-DAY RENEWAL VIEW contracts · health scores · risk flags surfaced automatically in CRM RENEWAL AS MOTION not an event that happens to you

"Renewals should not be a surprise conversation at month eleven."

In most companies, the renewal conversation starts three weeks before the contract anniversary — which is far too late to do anything about it. By then, if the relationship isn't in good shape, the best outcome is a last-minute negotiation that buys another year at a discount. The real conversation about whether to renew was had internally by the client months earlier, and nobody on your side was part of it.


The infrastructure I build treats renewal as a motion that starts at onboarding and runs continuously. Structured QBRs at the right intervals. A 90-day renewal pipeline view in HubSpot so you always know what's coming up and what the health of those accounts looks like. Renewal conversations that happen when the relationship is strong, not when the contract is expiring.

What I Build
01

Sales-to-CS Handoff Systems

Structured context transfer so CS inherits relationship intelligence — the stakeholder map, the promises made, the concerns raised — not just a deal record with three bullet points.

02

Onboarding Architecture

The 60-day window that determines whether an account renews is the most important commercial period of the entire relationship. Built accordingly.

03

HubSpot Lifecycle Stage Logic

Real stages with real automation — not labels that require manual updates nobody makes. The system reflects reality without human intervention at every step.

04

Health Scoring & Early Warning

Automated at-risk flags based on observable behaviour. The system surfaces accounts that need attention before they become churned accounts.

05

QBR & Cadence Infrastructure

Systematic attention on high-value accounts — not dependent on individual CS memory or calendar discipline.

06

Renewal Pipeline Visibility

90-day renewal view in HubSpot. Renewals treated as a motion, not an event that happens to you three weeks before the contract anniversary.

Retention starts with an honest look at your last ten churned accounts.
The diagnostic conversation is free. Let's look at what's actually there.
Expansion
03 · Expansion
Expansion treated as a motion. Not occasional luck.

Most companies treat upsell as something that happens when a client asks for more. Referrals are hoped for, not engineered. New markets are entered reactively. I build the systems that make expansion a deliberate, trackable, repeatable motion.

Upsell Infrastructure Referral Programs Market Entry NRR >100%
"Expansion revenue has 3× lower CAC than new logo acquisition. Most companies invest almost nothing in making it systematic — because it looks like it's working fine until it suddenly isn't."

The best pipeline you have is sitting in accounts you already own. Most companies leave it there — because there's no system to surface the right moment to ask, no structured referral program, and no new market entry playbook.

NRR above 100% is not a sales performance outcome. It's a systems design outcome.

The Problem
WHERE EXPANSION LIVES EXISTING ACCOUNT BASE accounts you already closed and retained UPSELL REFERRAL NEW MARKET client asks for more someone mentions you reactive entry no playbook ↑ these are not strategies CRM trigger at right time ask cadence tracked intro scoped first playbook ready NRR > 100%

"Upsell, referral, and new markets — all left to chance."

Expansion revenue has three times lower CAC than new logo acquisition. And yet most B2B companies invest almost nothing in systematising it. Upsell happens when a client happens to mention they need more. Referrals come in when a happy client happens to mention you to a colleague. New markets get entered when someone at a conference makes an introduction.


None of those are strategies. They're happy accidents — and happy accidents do not compound reliably enough to build a business on. The accounts you already own represent the most efficient pipeline available to you. The question is whether you have the systems to surface and convert it deliberately, or whether you're leaving it to whenever someone happens to ask.

UPSELL TRIGGER SYSTEM usage data contract terms health score EXPANSION TRIGGER right moment detection upsell play referral ask market entry TRACKED IN CRM EXPANSION PIPELINE deliberate · measurable · repeatable

"The right moment to ask is predictable. The system just needs to surface it."

An account that has been healthy for six months, whose usage has been climbing, and whose key contact just got a promotion is a very different expansion conversation than one where health is flat and the sponsor is disengaged. The signals that distinguish between them exist in the CRM — but only if the CRM was built to capture and combine them.


I build the trigger logic that identifies expansion moments before they pass. Usage thresholds that flag upsell readiness. Health and tenure criteria that determine when to ask for a referral. New market scoping that starts with data before committing budget. Each of these plays is tracked from initiation to outcome — so expansion stops being opportunistic and becomes a managed motion with a visible pipeline.

What I Build
01

Upsell & Cross-Sell Infrastructure

CRM-triggered, data-based expansion plays — not gut-feel asks at QBRs. The system identifies the moment and the motion, so the conversation happens at the right time.

02

New Market Entry Systems

Database scoping, targeting logic, and launch sequencing for new geographies. The work starts with scoping before any commitment is made.

03

Referral Program Design

Ask cadence, incentive framework, CRM tracking. Referral revenue engineered rather than hoped for — triggered at the right moment, tracked from the first introduction.

04

Channel & Reseller Partnerships

Agency partnerships, complementary providers, and reseller programs — with trackable pipeline attribution from day one.

05

New Revenue Stream Scoping

Adjacent services, product extensions, and untapped segments — assessed and sequenced before any investment is made.

The best pipeline you have is already inside your existing accounts.
Let's build the system that surfaces and converts it.
Partnership
04 · Partnership
Your brand.
My execution.
Zero operational burden.

For agencies, consultancies, and service businesses that keep getting asked for revenue operations capability they don't have in-house. Two models — white-label and co-delivery — depending on how involved you want to be.

"The agencies that offer RevOps without building it themselves are the ones adding the most revenue per client — without adding headcount, overhead, or delivery risk."
Zero operational burden 2 engagement models Your brand $48M+ pipeline built
Two Models
White-Label
Under your brand

I deliver the full engagement under your agency's name. Your client never knows I'm involved. You take the credit, invoice the client, and keep the margin. I handle all execution.

Co-Delivery
Shared engagement

We deliver together — your relationship, my revenue operations expertise. Structured for engagements where your team handles strategy and client management while I own the execution layer.

Who this is for
Agency Owners Growth Consultancies Sales Coaches & Trainers Marketing Agencies Independent Consultants
Interested in a partnership conversation?
Let's find the model that works for your business — no obligation on either side.
Here Are My Opinions

Revenue operations thinking — written from inside real engagements, not from the sidelines. Writing when there's something worth saying. Podcast when a topic needs more than a blog post can hold.

001
[No. 001]Acquisition

The ICP you built in year one is suppressing your pipeline in year three

Arushi Mittal·8 min read
002
[No. 002]Retention

Your HubSpot isn't a data problem. It's a workflow problem.

Arushi Mittal·6 min read
003
[No. 003]Retention

Churn is a handoff failure, not a CS failure

Arushi Mittal·9 min read
004
[No. 004]Acquisition

Why your pipeline stopped converting — and it's not the copy

Arushi Mittal·7 min read
005
[No. 005]Expansion

The referral revenue you're not asking for

Arushi Mittal·5 min read
006
[No. 006]Behind the Work

What I actually do in week one of a HubSpot engagement

Arushi Mittal·6 min read
[No. 001]Acquisition

The ICP you built in year one is suppressing your pipeline in year three

Arushi Mittal·8 min read·Acquisition

Most ICP documents were written once, during a period when the company needed to find its first customers, and they have never been meaningfully revisited. They sit in a Google Drive folder. They get referenced in onboarding decks. They inform the targeting logic that's been running — largely unchanged — since the sequences were first built.

And the reason pipeline generation feels harder than it should right now, in many cases, is that the ICP the team is targeting is no longer the buyer who actually closes. This is one of the most common and least discussed problems in B2B revenue operations.

How it happens

Year one ICP documents are built from a small number of data points: the founding team's network, the first handful of customers who said yes, and a set of assumptions about who the product is for that were reasonable at the time but haven't been tested against the reality of who's actually buying now.

By year three, the company has closed hundreds of deals. Some of those deals are healthy — good retention, expansion revenue, the kind of accounts the team would clone if they could. Some are not. High churn, difficult implementations, constant escalations, discount dependency. The mistake most teams make is treating this as a CS problem or a product-market fit problem. It's almost always a targeting problem. Those accounts were the wrong fit from the first conversation. The ICP document never got updated to reflect that distinction.

What targeting drift looks like in practice

The sequences are going to the same titles they always went to. VP Sales, Head of Revenue, CRO. The company size filter hasn't changed. The industry verticals are the same. Everything looks right on paper. But the data inside the CRM — if someone has looked at it carefully — tells a different story. The deals that close fastest, at the highest ACV, with the lowest discounting, with the best renewal rates, share specific characteristics that aren't reflected in the current targeting criteria. None of that is in the ICP document — because the ICP document was written before there was enough data to know any of it.

The fix isn't a targeting overhaul. It's a diagnostic.

The first thing I do in any acquisition engagement is run a closed-won analysis. Not a pipeline review — a closed-won analysis. Looking backwards at the deals that actually closed, what they had in common, where they came from, how long they took, and what the renewal looked like. That analysis almost always reveals a segment that is significantly outperforming the rest of the target universe. The question is never "how do we reach more people." It's "why are we spending 70% of our budget reaching people who look like the 30% of deals that cost us more than they return?"

What you can do this week

Pull your closed-won deals from the last 18 months. Filter by renewal rate, time-to-close, ACV, and number of support escalations in year one. Look at what the top quartile has in common — company size, industry, tech stack, hiring patterns, growth stage, org structure. That cluster is your real ICP. The question is whether your current targeting is pointed at it.

[No. 002]Retention

Your HubSpot isn't a data problem. It's a workflow problem.

Arushi Mittal·6 min read·Retention

When I'm brought in to look at a HubSpot instance that isn't working, the brief is almost always framed the same way: the data is a mess. Contacts haven't been updated. Lifecycle stages are wrong. Nobody knows what's actually in the pipeline. The instinct that follows is to clean the data. The instinct is wrong.

Data doesn't go bad on its own

The state of a company's HubSpot instance is a faithful reflection of the workflows — or absence of workflows — that surround it. If lifecycle stages are wrong, it's because there's no automated logic that moves contacts through stages based on behaviour. Someone is supposed to do it manually. Nobody does it manually, because manual data hygiene is not how any rep chooses to spend their time when there are conversations to have and deals to close.

Clean data in a broken workflow will be dirty data again within 90 days. Every time.

The four things I look at first

Stage logic. What are the criteria for moving a deal from one stage to the next? Are they based on rep judgment or on observable behaviour? Is there automation that enforces them or does movement depend on manual updates?

Handoff architecture. When a deal closes, what information moves to the CS team and in what format? Handoffs are where context dies — and context death is where HubSpot stops being useful.

Contact ownership. Who owns a contact after a deal closes? Ambiguous ownership is almost always the root cause of data that hasn't been touched in six months.

Automation coverage. The ratio of automated to manual updates is a direct predictor of data quality. If most updates require someone to remember to do them, most updates won't happen.

What a working HubSpot actually looks like

A CRM that does its job is not a place where data is stored. It's a system that reflects reality in near-real-time and surfaces the right information to the right person at the right moment without anyone having to go looking for it. That version of HubSpot exists. It's not a different tool — it's the same tool with the workflow architecture built properly underneath it. That's the work.

[No. 003]Retention

Churn is a handoff failure, not a CS failure

Arushi Mittal·9 min read·Retention

When retention numbers are bad, the conversation almost always ends up in the same place: what is the CS team doing, or not doing, that's causing accounts to leave? This is the wrong question. Not because CS doesn't matter — it does — but because in the majority of churn cases, the failure happened before CS was ever involved. It happened at the handoff.

What actually happens at the handoff

A deal closes. The AE has spent weeks, sometimes months, building a relationship with a set of stakeholders. They know which executive is the real champion and which one is a blocker. They know the internal politics around the budget approval. They know the main contact prefers Slack over email. None of that is in the CRM. What's in the CRM is: company name, deal value, close date, maybe a notes field with a few bullet points written at 11pm before a quarterly review.

The CS manager inherits this record and is expected to pick up the relationship as if they were present for every conversation. They're not. So they start from scratch. They ask the questions the AE already answered. The client — who was sold on a specific vision by a specific person — is now dealing with someone new who is rebuilding context that should have transferred automatically.

Why the 60-day window decides everything

The decision to renew is largely made in the first 60 days of an engagement, not in the weeks before the renewal date. By the time a CS manager is running a renewal conversation, the client has already decided — in some form — whether the engagement has been valuable enough to continue. If the handoff broke the relationship in week one, no amount of excellent CS work in month eleven will fully repair it.

What a functional handoff looks like

The handoff is not a meeting. It's not a warm introduction email. It's a structured transfer of commercial context: the specific outcome the client expects in the first 90 days, the stakeholder map, the promises made during the sales process, the concerns that were raised and how they were addressed, and what would make this client likely to expand. This information exists in the AE's head. The job is to extract it, structure it, and transfer it before the AE moves on to the next deal. HubSpot can be built to make this mandatory — not optional, not aspirational.

The conversation that needs to happen first

If your churn is above where it should be, before you look at CS processes, open the last ten churned accounts and look at the actual deal records at the point of close. What was there? What wasn't? How prepared was the CS manager for that first conversation? The answer is usually visible within five minutes. And it's almost never a CS problem.

[No. 004]Acquisition

Why your pipeline stopped converting — and it's not the copy

Arushi Mittal·7 min read·Acquisition

When conversion rates drop, the first instinct is to rewrite the sequence. Change the subject line. Shorten the message. Add more personalisation. These are not wrong things to do — but they are almost never why pipeline conversion stops working. The copy is a symptom. The targeting is the disease.

What actually changed

The sequences that worked twelve months ago were built for a specific buyer — at a specific stage of awareness, in a specific market context, dealing with a specific version of the problem your product solves. All three of those things change over time. Your buyer's awareness level shifts as your category matures. The market context changes as competitors enter and exit. The specific version of the problem evolves as companies grow through different stages. The sequence doesn't know any of that. It was written for a buyer that no longer exists in exactly that form — and it's being sent to a database that hasn't been meaningfully refreshed since it was first built.

Three things worth checking first

When did the database get built? If the core list hasn't been rebuilt with current data in the last six months, the problem may be as simple as sending accurate messages to people who've moved on, been promoted, or are no longer the right contact.

What is the sequence assuming about buyer awareness? A message written for someone who doesn't know they have the problem performs very differently for someone who's been evaluating solutions for three months. Most sequences don't adjust for that distinction.

Is the message addressing a problem the target actually has right now? Not a problem they had six months ago. A problem they are actively experiencing, visible through their hiring patterns, their recent announcements, or their company growth stage. Rewrite the copy last. It's usually the least leveraged change you can make.

[No. 005]Expansion

The referral revenue you're not asking for

Arushi Mittal·5 min read·Expansion

Most companies treat referral as something that happens when clients are happy enough to bring it up themselves. A satisfied client mentions you to a colleague. A champion moves to a new company and remembers you. This is not a referral program. It's luck with better branding.

What a referral motion actually requires

A referral program needs three things that most companies don't have: an ask cadence, an incentive framework, and CRM tracking that makes both visible and accountable. The ask cadence is the most important and most neglected. The moment a client is most likely to refer you is not at the twelve-month renewal. It's at the 60-day mark when they've just experienced the first clear win. It's the week after a QBR where the results looked good. It's immediately after a successful implementation. Those windows are predictable. Most companies don't build a system to capture them — they let them pass and hope the client remembers them later.

What needs to be built

The referral ask needs to be triggered automatically at the right moment — not dependent on a CS manager remembering to ask. The incentive structure needs to be simple enough that a client can explain it in one sentence. And the tracking needs to be in HubSpot so referral-sourced pipeline is visible, attributable, and comparable to every other acquisition channel. None of this is technically complex. Almost no company has built it properly.

[No. 006]Behind the Work

What I actually do in week one of a HubSpot engagement

Arushi Mittal·6 min read·Behind the Work

The most common expectation when an engagement starts is that week one involves setting up tools, aligning on timelines, and beginning the work. In a HubSpot engagement, week one is almost entirely diagnostic — and what it reveals usually changes the scope of everything that follows.

The audit before the proposal

Before I make a single recommendation, I want to understand four things: how stage logic actually works versus how it's supposed to work, where the automation is and where it's conspicuously absent, what happened at the last five handoffs and what information transferred, and what the data looks like on the accounts that churned in the last 12 months versus the ones that renewed and expanded. That last question is the one most engagements skip. Churned accounts are the most honest data in any CRM. They show you where the system failed — at handoff, at onboarding, at the point when someone should have flagged an at-risk signal and didn't — in a way that no amount of pipeline reporting can.

What week one usually reveals

In most HubSpot instances that have been running for more than a year, week one reveals some version of the same pattern: stage definitions that no longer reflect the real sales process, automation that was set up once and never maintained, and handoff documentation that doesn't exist in any systematic form. The good news is that none of this is irreversible. The bad news is that cleaning it up properly takes longer than anyone wants to hear in week one — which is why the diagnostic has to come before the timeline, not after it.

Why the diagnostic changes everything downstream

Every proposal I write is based on what the diagnostic finds, not on what a typical engagement looks like. Some HubSpot instances need a complete post-sale rebuild. Others need surgical workflow fixes and a new handoff process. Some need the ICP looked at before any CRM work makes sense. Week one is the work that makes everything else worth doing. Without it, you're optimising a system you don't actually understand yet.