Tayyub Yaqoob
Contact →
Home›Case Studies

Case studies

Five projects in production. Each card has the stack, the headline numbers, and (where appropriate) a note from the team it shipped to.

Forecasting · BigQuery · GA4

Channel and query forecasting

Forecasting workflows in BigQuery covering GSC query trends, channel traffic, and monthly reporting cycles. Analyst and account teams use them to catch variance mid-month; confidence bands sit alongside the point estimates so the range is visible.

120k 90k 60k NOW Jan Jun Dec
BigQueryPythonGSC APIGA4ForecastingReporting
Median MAPE5.4%
Runs / month12
Channels4

"The tool is working fully, and we've been able to share the update with wider teams."

BBC, project stakeholder
Sentiment · LLM-assisted

Sentiment and topic tracking

LLM-assisted classification across press, reviews, and social. The system classifies, deduplicates, and flags shifts — output queues into an analyst review before anything leaves the room. Annotated bands show positive, neutral, and negative coverage over time.

Coverage spike · Q2 Topic shift · Q4 Jan Jun Dec positive neutral negative
PythonLLM APIBigQueryNLPStructured outputs
Surfaces tracked11
Mentions / week~3k
Alert classes4

"The feedback was very positive. Overall they are happy with the first go at this. I appreciate all your help and of course passed on high praise of your work."

General Mills, account team on Blue Buffalo
AI visibility · AI Overviews

AI vs traditional search

Weekly tracking of how brands and queries surface across AI Overviews, ChatGPT, Perplexity, and regular Google. Each snapshot stores the date, prompt, and rendered citation so visibility shifts stay reviewable months later.

Google · organic competitor a · category guide competitor b · review hub your brand · category page forum · long-tail thread AI Overviews · cited your brand · cited 1/3 competitor a · cited 2/3 trade outlet · cited 3/3 prompt: "best [category] tool · 2026" Jan Jun Dec
AI OverviewsChatGPTPerplexitySERPResearch
Queries tracked140+
Platforms4
CadenceWeekly
Reporting automation

Automated report assembly

Monthly reports and decks pulled from GA4, GSC, and ad-platform data and rendered into Google Slides. What used to be a half-day manual build now lands as a 20-minute analyst review. The bar chart maps the real stages and what each one used to cost.

0h 1h 2h 3h 4h Manual Gather ~60 min Clean ~50 min Build deck ~60 min Review ~40 min Send ~30 min 4h · manual vs Automated Review ~20 min 20m · review
PythonGoogle Slides APIBigQueryAutomationClient delivery
Reports / month36
Time saved~70%
Variance flagsInline

"Our analytics are in great shape at this point and we're feeling confident about the ways they will help us."

Diligent Pharma, stakeholder
Machine learning · embeddings · NLP

Competitor content mapping

Embeds client and competitor pages with a sentence-transformer model, then uses cosine similarity to rank the closest competitor matches for each client URL. Analyst teams use it to spot direct content overlap and prioritise pages that need a response.

Competitor A Competitor B Competitor C Competitor D Competitor E Competitor F Client /pricing Client /features Client /compare Client /docs Client /blog .71 .42 .89 .21 .38 .17 .46 .78 .41 .91 .62 .24 .19 .66 .44 .37 .86 .28 .39 .45 .59 .22 .41 .83 .81 .69 .26 .43 .36 .31
Pythonsentence-transformersBeautifulSoupNumPyCosine similarity
Pages embedded500+
Top-k matches5
Embedding dim768

Want a walkthrough of any of these? Email me.