Skip to main content
MethodScience & Data AnalysisData analysisAdvanced

Build a simple time-series forecast

A baseline forecast with method comparison, backtest error and prediction intervals.

This is a practical, copy-paste process for projecting sales, prices, traffic or demand.. Use it when you want a clear structure, then add your own details and the linked evidence.

Published by Deepstory Research. Reviewed Sep 2026.

How to use this method

  1. 1. Copy or downloadCopy the prompt below, or download the method file and attach it to your chat.
  2. 2. Paste into any AIChatGPT, Claude, Gemini or Copilot. Replace each [PLACEHOLDER] with your details, or let the AI ask you.
  3. 3. Add evidenceAttach the evidence files listed on this page so the AI works from sourced numbers, not guesses.

The prompt

Paste it as-is. The AI will ask for anything still in [BRACKETS].

Download .md
# Build a simple time-series forecast
Method by Deepstory Research: deepstoryresearch.com/method/time-series-forecast

You are a forecasting analyst. Follow the method below step by step. Goal: A baseline forecast with method comparison, backtest error and prediction intervals.

## My inputs
- Series: [SERIES]  (example: attach monthly data)
- Horizon: [HORIZON]  (example: 12 months)
- Tool: [TOOL]  (example: Python, Excel)

Before you start: if any input above is still in [BRACKETS] or missing, ask me for it in one short numbered list and wait. Do not guess. If I attached files (data, briefs, Deepstory evidence files), treat them as the primary source.

## Steps to follow
1. Plot and decompose trend and seasonality.
2. Fit naive, seasonal naive and exponential smoothing baselines.
3. Backtest and compare error.
4. Forecast with intervals.
5. Explain the drivers that could break the forecast.

## Deliver the output in this structure
- Decomposition
- Model comparison
- Forecast table with intervals
- Code
- Risks

## Quality rules
- Beat the naive baseline or say the model adds nothing.
- Label every number as FACT (with source and date), ESTIMATE (show the calculation) or ASSUMPTION.
- If a Deepstory evidence file is attached, use its figures as the reference values and cite its source references; tag claims that are not from it as [UNVERIFIED].
- End with: the 3 assumptions that matter most, what would change the answer, and the single best next step.

What you need to give the AI

InputPlaceholderExample
Series[SERIES]attach monthly data
Horizon[HORIZON]12 months
Tool[TOOL]Python, Excel

Use it when

  • Projecting sales, prices, traffic or demand.

What you get

  • Decomposition
  • Model comparison
  • Forecast table with intervals
  • Code
  • Risks

The procedure, step by step

Read this before you run it, so you can check the AI actually followed the method.

  1. 1

    Plot and decompose trend and seasonality.

  2. 2

    Fit naive, seasonal naive and exponential smoothing baselines.

  3. 3

    Backtest and compare error.

  4. 4

    Forecast with intervals.

  5. 5

    Explain the drivers that could break the forecast.

Quality checks built into the prompt

  • • Beat the naive baseline or say the model adds nothing.
  • • Every number labelled FACT, ESTIMATE or ASSUMPTION; claims outside attached evidence tagged [UNVERIFIED].

Evidence to use with this method

Methods tell the AI how to think. Evidence gives it sourced numbers to think with. Download a context file and attach it with the prompt.

EvidenceEnergy

US Oil & Gas Dashboard

Every key US oil & gas number on one page: the 2026 crude-price shock from the Strait of Hormuz, Henry Hub natural gas, record US crude production, a depleted Strategic Petroleum Reserve, the drilling rig count, and what it all means at the pump.

Open evidenceDownload context file
EvidenceFinance

US Inflation & the Fed

The full 2026 inflation story on one page: the CPI path and category breakdown, the mid-year turn and the still-rising core PCE outlook, and the Fed’s rate response — fed funds, the 10-year Treasury, and the 30-year mortgage.

Open evidenceDownload context file
EvidenceGeopolitical

Internet use ranking

Which selected economies reported the highest internet-use rates? Individuals using the Internet (% of population) · common year 2024 · 20 selected economies with published values

Open evidence

Run next

Methods that chain well after this one.

Operations, Supply Chain & ManufacturingForecast product demandA demand forecast with method choice, seasonality, accuracy measure and a planning range.Science & Data AnalysisClean a messy datasetA data-quality report and step-by-step cleaning plan (with code) that keeps an audit trail.Science & Data AnalysisExplore a dataset (EDA)A structured exploration: distributions, trends, segments, relationships and questions worth testing.Science & Data AnalysisChoose and design the right chartThe chart type that answers your question, with a design spec that avoids common distortions.

Deepstory methods are process guidance for AI tools. Review AI output before relying on it; legal, tax, medical and investment decisions need a qualified professional.