How accurate are the revenue projections?
Our projections use established statistical methods including exponential smoothing, linear regression, and seasonality decomposition. Accuracy depends on data quality and consistency — we recommend at least 6 months of clean historical data. Every projection includes confidence ranges so you can see the likely margin of error. Projections within 30 days are typically more reliable than 90-day ones, which is why we present three scenarios.
What is the difference between optimistic, base, and conservative scenarios?
The optimistic scenario assumes your best recent growth trends continue. The base scenario uses weighted averages of historical patterns including seasonality. The conservative scenario applies lower growth assumptions and factors in potential downside risks. Together, these three projections give you a realistic range of likely outcomes rather than a single misleading number.
How long does it take to receive my forecast?
Most forecasts are delivered within 2–5 minutes of payment confirmation. Larger files with complex data may take up to 10 minutes. You will receive an email with your PDF forecast attached.
What happens to my uploaded file?
Your file is encrypted and stored temporarily in Cloudflare R2. Once your forecast is generated and delivered, the raw file is automatically deleted within 24 hours. The generated PDF forecast is kept for 30 days so you can re-download it if needed.
What format should my data be in?
We accept CSV files, Excel files (.xlsx and .xls), and public Google Sheets links. Your data should have columns for date and revenue amount at minimum. The more history and context you include, the better the forecast quality.
Does anyone at Tenor Analytics read my data?
No. The entire forecasting pipeline is automated. No human at Tenor Analytics reads, reviews, or accesses your uploaded data. Your file goes into our system, Python processes it automatically, the forecast is generated, and the file is deleted. Only aggregated statistical figures are used — never your raw data.