You bought the analytics tool. Now get the decisions.
Liftpoint Analytics is a fractional Head of Product Analytics & ML: a senior analytics lead two days a week, scoped as fixed packages with a named decision at the end of each one.
For B2B SaaS and marketplace companies at $5–75M ARR that run Fullstory, Pendo, Amplitude, Mixpanel or GA4 and still make big calls on anecdotes.
Three things the role covers
Each one has a number behind it from prior work at a publicly traded two-sided marketplace for high-value physical goods.
-
Experimentation at scale
Design the test, read it honestly, and get the decision signed off by the executive who asked for it.
30+ A/B tests run. CUPED variance reduction cut test durations by about 25%. A listing-page redesign shipped to 100% of users after the exit-rate scare turned out to be a funnel shift.
-
Instrumentation ownership
Own the tracking plan, the event taxonomy and the vendor relationship, so the data is trustworthy before anyone reads it.
GA4 tag-manager tagging was losing an estimated 15–20% of data. Replaced with a digital-experience platform in about 6 months, adopted by 9+ teams. Issue resolution went from about 3 days to 1.
-
ML and economics that pay for themselves
One decision, one model, a dollar outcome someone can audit.
Break-even modeling on a delivery pilot moved programmed margin from 5% to 15% and take rate from about 0.4% to about 3% of eligible listings in about 14 months. Gradient-boosting pricing work tied to $1.2M in annual savings.
If any of these sound familiar
These are the sentences product leaders say in the first ten minutes of a call.
- “We pay for Fullstory and I'm not sure anyone opens it.”
- “Every A/B test ends in an argument about whether it worked.”
- “The CEO asked for a number and three dashboards gave three answers.”
- “Our analyst is great at SQL but can't present to the exec team.”
- “We can't justify a $180K analytics hire yet, but we're flying blind.”
- “We know pricing is leaving money on the table. We don't know how much.”
Three ways to work together
Fixed fees or monthly retainers, not open-ended hourly billing. Start with the Diagnostic unless you already know which decision you need.
-
Offer 1 · Start here
Analytics Diagnostic
$6,500 fixed
2 weeks. Credited against A or B if signed within 30 days.
Two weeks of read-only access and interviews. Where the data is, what tracking is broken, which dashboards matter, how decisions actually get made, and the top three decisions a model could improve.
- Up to 6 stakeholder interviews
- Tracking health check on up to 20 core events
- Maturity score, 90-day roadmap, written report and a 60-minute readout
-
Offer A
Fractional Head of Product Analytics
from $9,000 per month
Two days a week. 3-month minimum, then month to month. $9,000–12,000.
A senior analytics lead who owns the roadmap, designs and reads tests, builds executive reporting, mentors your analyst and sits in product leadership meetings.
- Tracking plan and event taxonomy owned, not outsourced
- Test calendar with a decision log for every test
- Monthly readout you can forward to the board
-
Offer B
ML Decision Model
from $30,000 fixed
6–10 weeks. $30,000–50,000 fixed, or base plus a share of measured lift.
One decision, one model: pricing, segmentation, churn, demand forecasting or delivery and ops economics. Scoped around a dollar outcome and a break-even.
- Data assessment, model, validation and deployment plan
- Plain-English readout the sponsor has already seen before the meeting
- Lookback after launch to check the model held in the field
Pricing within a range depends on the number of product teams, data sources, whether a warehouse exists and whether there is an analyst to mentor. Net-15 terms on every invoice.
Need something smaller?
Fixed-price foundations work, from $1,500.
A dashboard, a tracking audit, a tracking plan, one A/B test read properly, a KPI workshop, a segmentation or a forecast. Fixed scope, fixed fee, and each one says plainly what is not included.
Three outcomes from past work
Prior employment, anonymized. Every number was reported to the executive who asked for it. All case studies
About
US-based, remote.
Liftpoint Analytics is run by a product analytics lead who works with companies that have the data but not the senior person to turn it into decisions. Five-plus years in analytics and data science; the last two-plus running a product analytics pod for the Chief Product Officer, then the Chief Marketing & Analytics Officer, of a publicly traded two-sided marketplace for high-value physical goods. Before that, analytics at a large transportation infrastructure operator, where ML forecasting improved forecast accuracy about 10%.
How the work runs. Every engagement names the decision it supports, the owner and the date it is needed; analysis that does not change a decision is cut. The tracking plan, event taxonomy and QA come before dashboards, and dashboards come before tests. Every test gets a hypothesis, a powered design, guardrail metrics and a decision log. Every readout is one page a CPO can forward.
I will tell you what the data cannot answer before I tell you what it can.
- Certifications
- AWS Certified Machine Learning – Specialty
- Tools
- SQL, Python (pandas, scikit-learn, XGBoost, PyTorch), BigQuery, Snowflake, dbt, Tableau, Power BI, Fullstory, GA4, Pendo, Statsig
- Capacity
- Max three retainer clients at once.
- Code
- Public ML demo on GitHub [TODO: GitHub URL]
Book a free 30-minute Decision Audit
Ten questions, three observations, live. You leave with a recommendation: the Diagnostic, nothing yet, or not a fit. No slides.
Prefer to pick a slot now?
Book directly on the calendar. The form above is optional if you do.
A decision, not a dashboard.
Two days a week, three-month minimum. You get a tracking plan, a test calendar and a monthly readout you can forward to the board.