Performance and growth marketer. I help companies get more customers who actually pay, from paid ads, search and now AI answers. Ten years in, from my own e-commerce stores to $20M+ a year in paid at Shopify.
Short version: I’m an independent growth and performance marketing consultant for B2B SaaS and ecommerce companies. I find the money you’re spending that isn’t bringing in paying customers, move it to what is, and set your team up to keep it going without me. I’ve done it across $20M+ a year in paid at Shopify, at Pipe17 where qualified pipeline got 74% cheaper, and at a ten-practice dental group where new patients got 56% cheaper.
How I can help.
Three things I do for teams. Most projects start with a look at your accounts and data and a written plan. Then I build it with your team, or for them.
I audit your accounts against the customers who actually pay, not platform conversions, test what’s incremental with holdouts and move budget to what works. As a project, or as a fractional growth lead while you hire.
What you get
An audit tied to revenue
Bidding that learns from paying customers
A holdout test you can rerun
Geo holdouts at Shopify found $2M+ in paid social that wasn’t adding anything. At Pipe17, cost per qualified opportunity fell 74%.
I build a fixed panel of the questions your buyers ask ChatGPT, Perplexity, Gemini and Google’s AI Overviews, measure who gets named and cited, then build the pages and mentions that get you in. Then I connect it to your CRM so you can see what it’s worth.
What you get
Your citation share next to competitors
A ranked list of pages, with the first ones written
AI referral revenue tracked in GA4 and your CRM
At Pipe17 this found $960K in closed-won revenue from AI-referred buyers, 32% of new-business wins.
I set your marketing team up to run reporting, research, ad ops and page builds on Claude Code or Codex, connected to your real accounts, then show them how to do it themselves. A person still approves anything that spends money.
What you get
A working setup on your own accounts
Skills for the jobs your team repeats
Hands-on training, so the team runs it next week
At Pipe17 a 12-agent research pipeline grew the target list from 441 to 2,795 accounts.
CommasHead of Growth MarketingMy current role, so the details stay private.
Rates.caPerformance Marketing ManagerPaid acquisition for an insurance comparison marketplace in Canada, and landing page tests with product and engineering from ad click to completed quote. I also wrote the team's A/B testing overview.
The setup I run every day: AI agents connected to your ad accounts, CRM and analytics, with skills for the work your team repeats and a person approving anything that spends money.
Your systems
GA4MCP
Google AdsCLI
Meta AdsAPI
HubSpotCLI + MCP
SalesforceMCP
Search ConsoleAPI
LinkedIn AdsMCP
Docs and notesSync
Your teamPriorities, judgment and every spend approval
Research agentAnalysis, research and browser tasks, done with your team
Long-running agentScheduled and overnight work that runs on its own
Rules and factsWhat needs approval, which tools to use, and the account facts agents read first so they never guess.
SkillsThe weekly audit, the report, the research brief. Written once, reused every week.
Scheduled jobsReports and alerts that run on their own and only ping someone when a decision is needed
What your team gets
Paid media opsDaily
SEO and AEOWeekly
AttributionWeekly
LifecycleOngoing
Account researchOn demand
ReportingScheduled
Sites and funnelsOn demand
CreativeOn demand
Hover over any piece to see how it connects. Nothing that spends money goes out without a person’s OK.
6.3xbigger target account list from a 12-agent research pipeline
13,000+companies classified for tiering by agents
$24Ka year in software replaced by tools I built
I wrote up the whole setup, including how a marketing team could start doing this next week.
If your brand isn’t in that answer, you weren’t considered, and nothing in your analytics tells you. I measure where you stand and build what gets you in.
Pipe17 · Attribution
$960Kclosed-won
came from buyers an AI assistant sent.
That was 32% of new-business wins, and the CRM had credited it to search. Asking buyers what they actually used, and tracking referrals from ChatGPT, Perplexity, Gemini and Claude every month, put it where it belonged.
AI-referred wins32%
How I measure it
7
numbers, tracked every week.
Mention rate, citation share, answer position, source mix, answer accuracy, AI referral sessions and what buyers say in your how-did-you-hear field. All from a fixed panel of the questions your buyers actually ask.
I built this site the same way: clear facts about who I am, structured data, an llms.txt file and AI crawlers allowed in. If an AI assistant sent you here, it worked.
I was the e-commerce kid. I sold iPhones and Xboxes at school, flipped domains, then built my own online stores, including a hoodie store at McMaster, on Facebook ads and Instagram influencers I figured out from YouTube. A few outlets wrote about it back in 2020. That turned into a small agency, then I did the same job at much bigger scale, from a traditional shipping company I talked into paid digital to three years running paid at Shopify.
I’m Canadian, I live in Miami, and today I’m Head of Growth Marketing at Commas. The question hasn’t changed in ten years: which dollars bring in customers who actually pay? I like working with teams who want that answer and want to keep finding it after I leave.
“The engine room of our growth motion… it’s moved the numbers. Our lead-to-SQO conversion and cost-per-influenced-SQL both prove it.”
VP Marketing, Pipe17
Quick answers about me.
Who is Farhan Munshi?
Farhan Munshi is a performance and growth marketing consultant based in Miami and Head of Growth Marketing at Commas. He started out in e-commerce, building and marketing his own online stores, and has since run $20M+ a year in paid media at Shopify and growth programs at Pipe17, Dental Team FL and Purolator.
What does he help companies with?
Performance marketing and growth first: making paid acquisition on Google, Meta, TikTok and LinkedIn pay by optimizing to customers who actually pay and proving it with holdout tests, often as a fractional growth lead. He also gets brands named in AI answers like ChatGPT (AEO), and sets marketing teams up to run their work on AI agents.
Does he take on consulting or freelance projects?
Yes. He takes on a small number of projects: AEO programs, paid acquisition audits and rebuilds, and AI agent setups and training for marketing teams. The fastest way to start is an email to [email protected].
What has he done in AI search (AEO)?
At Pipe17 he rebuilt attribution around what buyers said they used and found that AI-referred buyers accounted for $960K in closed-won revenue, 32% of new-business wins, which the CRM had credited to search. His playbook is published at farhanmunshi.co/notes/aeo-playbook.
What results has he delivered?
Cost per sales-qualified opportunity down 74% at Pipe17, Shopify paid media scaled 4x at a 3.2x ROAS with geo holdouts that found $2M+ in non-incremental spend, and cost per new patient down 56% across ten dental practices with no added spend.
How does he use AI agents in marketing?
Claude Code and Codex run his reporting, research, ad operations and page builds, connected to his own tools for Google Ads, HubSpot, UTMs and enrichment, with a person approving anything that spends money. He sets the same system up for marketing teams and trains them to run it.