Explore Our Services
Explore our range of services designed to help you move forward with confidence, wherever you're headed next.
What We Offer
From first assessment to forecasts that run on their own.
Every organization makes decisions based on what it expects to happen: how much power to buy, what to build, which policy to adopt, how much inventory to carry. User Metrics makes those expectations more accurate and shows how much confidence each one deserves.
Our services cover the full path. We assess the forecasts you already rely on, build better ones, model the decisions in front of you, independently review numbers prepared by others, and deliver the software that keeps it all running. Most clients start with a Forecast Audit and grow from there.
Every engagement is led by senior expertise from start to finish, and every model comes with plain-language documentation your team can understand and defend.
-
Find out how good your forecasts really are.
A fixed-price assessment of your current forecasts against actual results. We measure accuracy, find where and why misses happen, quantify which errors cost the most, and test a benchmark model on your data.
You get: an accuracy report, a cost-of-error analysis, an interactive dashboard and a team walkthrough.
Typical timeline: two to three weeks. -
Know what's likely. Plan for what's possible.
Custom forecasting models for load, demand, renewable generation, sales or inventory. Every forecast includes the likely outcome, the range of possibilities and the probability of crossing the thresholds that matter to you.
You get: production-ready models, forecasts at the horizons you plan on, full documentation and a handoff your team can maintain. -
See the consequences before you commit.
Models that show what happens under different policies, technologies or market conditions, such as electrification mandates, EV adoption, rate changes or extreme weather. Each scenario comes with clear assumptions and quantified uncertainty.
You get: scenario results, decision-ready briefings and models you can rerun as conditions change. -
A second opinion you can defend.
An objective assessment of forecasts and models prepared by others, such as utility load forecasts, consultant projections or vendor models. Built for agencies, advocates and decision makers who need to know whether the numbers hold up.
You get: a written assessment of methods, assumptions and risks, plus alternative estimates where warranted. -
Put forecasts in front of the people who decide.
We build the software that makes forecasts usable: automated data pipelines, forecast engines and interactive dashboards. They update on their own and fit how your team actually works.
You get: deployed software, source code and documentation, with optional ongoing support.
Our Process
Discovery
We learn the decision behind the forecast: who uses it, what's at stake and where current numbers have fallen short. Then we review your data and define what success looks like.
Model Build
We build and test forecasting models against your own history, so you see how they would have performed before you rely on them.
Delivery
You get the forecasts, the model and a plain-language report covering assumptions, accuracy and risks. We walk your team through it until they can explain it themselves.
Ongoing Support
Conditions change. We offer monitoring, regular forecast updates and model retraining as needed, monthly or quarterly.
Featured Work
San Diego Electricity Demand Outlook
Probabilistic Energy Demand Forecast
The question: How much electricity will San Diego need over the next 12 months, and how high could peak demand go during extreme heat?
The approach: We built a probabilistic 12-month outlook from public hourly demand data from the California Independent System Operator, combined with 30 years of San Diego weather records. A Bayesian model learns how demand responds to heat, cold, the calendar and long-term growth. We then simulated the coming year thousands of times, once for every weather year on record, so the forecast reflects both what the weather might do and how uncertain growth is. Before trusting it, we tested the method on the most recent 12 months.
What it shows:
Expected monthly electricity use, with the range of plausible outcomes
The "1-in-2" and "1-in-10" annual peaks that planners design around
The probability that peak demand tops [X] megawatts, such as San Diego's all-time record of 5,032 MW
How much of the uncertainty comes from weather versus underlying growth
The months when the annual peak is most likely to land
**An independent analysis using public data. Not affiliated with San Diego Gas & Electric or the California Independent System Operator.