Skip to content

taha.

Back to home taha.

✦ Selected work

Case studies

A marketing agency with no outbound channel. A cleaning company whose market ran dry. A staffing firm paying to enrich people it was never going to contact. Fifteen engagements: what was broken, what I built, and what it returned.

Client engagement

Outbound in Norwegian

A 25.26% reply rate across 18,062 Norwegian prospects, in a language I do not speak.

25.26%Reply rate

68.9%Best campaign, 73 of 106 replied

18,062Prospects · 14 months

SmartleadApolloGoogle Sheets

Read full case study

Client engagement

The Enrichment Engine

A recruiting engine rebuilt to decide before it pays for data, then pointed at two more verticals.

3Verticals on one engine

~50,000Suppression list

4Qualification gates

ClayLemlistRecruiterFlowAirtableApify

Read full case study

Client engagement

Outbound in Norwegian

A 25.26% reply rate across 18,062 Norwegian prospects, in a language I do not speak.

25.26%Reply rate

68.9%Best campaign, 73 of 106 replied

SmartleadApolloGoogle Sheets

Read case study

Client engagement

The Enrichment Engine

A recruiting engine rebuilt to decide before it pays for data, then pointed at two more verticals.

3Verticals on one engine

~50,000Suppression list

ClayLemlistRecruiterFlowAirtableApify

Read case study

Client engagement

The Listing Auditor

A pipeline that audits a brand's live Amazon listing and writes the email about what it found.

11Listing checks per brand

44Positive replies, one campaign

ClayOpenAISmartleadMarketplace data

Read case study

Client engagement

Rebuilding a Dry Market

The addressable market ran out, so I scraped 70,000 records and built a new one.

70,000Records scraped

1,407New companies found

Google MapsClayApolloSmartlead

Read case study

Client engagement

Compliance, Sold Cold

23 campaigns selling anti-money-laundering software to the compliance officers who buy it.

23Campaigns · 18 months

InstantlySmartleadClayApolloLinkedIn Sales Navigator

Read case study

Client engagement

One System, Two Payment Rails

Two payments platforms sold to regulated finance across Europe on one shared system.

7.3%Payments reply rate

7.0%Gateway reply rate

SmartleadClayApolloLinkedIn Sales Navigator

Read case study

Built and running

The Podcast Pipeline

Finds founders interviewed on small podcasts, reads the episode, and writes about what they said.

5,619Episodes mined

867Prospects delivered

PythonasyncioPostgreSQLOpenAISerper

Read case study

Client engagement

Creator Content, Gulf-Wide

Arabic creator content sold across the Gulf, where offering something beat asking for something.

13.2%Reply rate

29.8%Best campaign

SmartleadApolloGoogle Sheets

Read case study

Client engagement

Consulting, Sold Cold

The first client, the hardest category, and the engagement where the whole machine was learned.

116Companies interested

90,981Emails sent

ClayApolloSmartleadGoogle Maps

Read case study

Client engagement

New Market, First Try

On-site workplace fitness sold to Oslo employers in Norwegian: 20.7% on the first campaign into a brand-new market.

20.7%First campaign reply rate

4,411Prospects contacted

SmartleadClayApollo

Read case study

Client engagement

The Market Decides

Same product, same email: 0.9% in the US and 2.8% in MENA. Market selection, measured.

2.8%MENA reply rate

0.9%Same email, US

SmartleadApolloGoogle Sheets

Read case study

Client engagement

Vertical Fit, Measured

Same offer, same copy: education returned fourteen times the interested replies of law firms. The list is the strategy.

14xEducation vs legal interest

14Interested · education

SmartleadApolloGoogle Sheets

Read case study

Client engagement

The Trust Purchase

LinkedIn ghostwriting sold to founders, opened with a free audit of their own profile.

44Positive replies

11,799Founders reached

SmartleadApolloGoogle Sheets

Read case study

Built and running

The Review Engine

4,809 public reviews mined, scored by aspect, and written into a boardroom deck by the system itself.

4,809Reviews mined

6Aspects scored

PythonpandasOpenAIpython-pptx

Read case study

Built and running

The Model Router

Two trained classifiers behind an LLM agent that decides which one to call.

92.4%Risk model accuracy

98.4%Spam model accuracy

Pythonscikit-learnLangGraphStreamlit

Read case study

© 2026 Inam Taha

LinkedIn Email Top ↑