If you’re tired of “analytics” tools that promise the moon but leave you staring at confusing dashboards, you’re not alone. This guide is for anyone who actually needs to forecast sales revenue—owners, ops folks, sales managers—and wants clear, realistic steps to get Empler analytics doing useful work for you. No magic wands, no buzzwords. Just what you need to know, and what you can skip.
1. Get Your Data Ready (Don’t Skip This)
Before you even touch Empler, make sure you have the right sales data handy. Empler’s forecasts are only as good as what you feed it, so don’t expect miracles from garbage-in.
What you’ll need: - Historical sales data (minimum 12 months, ideally 24+) - Consistent time intervals (monthly, weekly, or daily—pick what matches your sales cycles) - Revenue numbers (not just units sold) - Any key factors that influence your sales (promos, seasonality, product launches)
Pro tips: - Use a spreadsheet to clean up your data before importing. Remove duplicates, fix typos, and fill in missing values if you can. - If your data is scattered across different tools (CRM, spreadsheets, billing), consolidate it first. Don’t try to “fix it later”—you’ll regret it.
2. Set Up Your Empler Account
If you’re new to Empler, create an account and get familiar with the interface. Empler’s not the most beautiful tool on the planet, but it’s straightforward once you get used to it.
Steps: 1. Sign up and verify your email. 2. Log in and poke around—see where your datasets, dashboards, and forecasting tools live. 3. Review the pricing and data limits. Some features might be locked behind a paywall. Decide if you can live with the free tier or need to upgrade.
What to ignore:
All the “AI-powered insight” banners. Focus on learning where to import data and run forecasts. The bells and whistles can come later.
3. Import Your Sales Data
Now, bring your cleaned sales data into Empler.
How to import: - Go to the Data section (sometimes called “Datasets”). - Click “Import Data.” - Choose your format (CSV is safest—Excel sometimes causes hiccups). - Map your columns clearly: Date, Revenue, and any Extra Factors (promotions, product category, etc.).
Watch out for: - Date formats that don’t match (Empler likes YYYY-MM-DD best). - Columns that get merged or dropped—double-check the preview before confirming. - If you’re connecting a live source (like Salesforce or Shopify), follow Empler’s guides. But know that integrations can break; a manual CSV is more reliable at first.
Pro tip:
If your sales data has gaps (weeks with no sales), fill in with zeros instead of leaving them blank. Empler treats blanks as missing, not zero.
4. Configure Your Forecast Model
Here’s where most people get stuck: Empler offers a handful of forecasting models, and the defaults aren’t always the best for you.
Basic steps: 1. Go to the Forecast section. 2. Select your imported dataset. 3. Choose your target column (usually “Revenue”). 4. Pick a time interval (weekly, monthly, etc.).
Model choices:
Empler typically gives you options like:
- Simple Moving Average — Quick and dirty, fine for steady sales, but useless if you have seasonality or big swings.
- Exponential Smoothing — Handles trends a bit better, but can lag behind rapid changes.
- ARIMA/Prophet — Handles seasonality and more complex patterns, but takes longer to set up and is more likely to throw weird errors if your data’s messy.
What actually works:
- Start with Exponential Smoothing if you’re new. It’s forgiving and catches basic trends.
- If you’ve got lots of seasonality (holidays, back-to-school bumps, etc.), try Prophet or ARIMA, but only after your data is squeaky clean.
- Ignore the temptation to toggle every “Advanced” setting unless you know what you’re doing. More knobs usually means more ways to mess up.
Pro tip:
Run the forecast on a subset of your data (e.g., the last year) and compare the output to what actually happened. If it’s wildly off, your model or your data needs work.
5. Fine-Tune Your Forecast (Optional, But Worth It)
Empler lets you tweak model parameters—things like seasonality, trend smoothing, and outlier handling. Here’s what’s actually worth your time:
- Seasonality: Turn it on if you know your sales cycle (e.g., retail spikes in December). Set the season length (e.g., 12 months or 52 weeks).
- Holidays/Promos: If you have a list of known promotional dates, import them as external factors. Empler can adjust its forecasts accordingly.
- Outlier removal: If you had one-off events (a huge client ordered once and disappeared), mark or remove those to avoid skewing your forecast.
What to ignore:
Most “Auto-tune” buttons. They sound smart, but they often overfit your data, making your forecast look perfect—until it isn’t.
Pro tip:
Keep notes on what you change. If the forecast gets worse, roll back your tweaks.
6. Review and Interpret the Forecast
Once Empler spits out a forecast, look at it with a skeptical eye.
What to check: - Does it pass the sniff test? If it predicts a 300% sales jump out of nowhere, something’s off. - Forecast intervals: Empler will show a “confidence interval” (the shaded area). If it’s huge, your data is noisy or inconsistent. Don’t trust the numbers too much. - Compare to reality: Overlay last year’s actuals. If the forecast completely misses known patterns, fix your data or try a simpler model.
Red flags: - Forecasts that are too smooth (no ups and downs) are probably ignoring seasonality. - Spiky, jumpy forecasts mean your data is messy or the model is overreacting.
Pro tip:
Forecasts are just a guide, not gospel. Use them to inform decisions, not make them for you.
7. Set Up Dashboards and Alerts
Once you’ve got a forecast you trust (or at least don’t hate), turn it into something actionable.
Steps: - Create a dashboard with your forecast, actuals, and confidence intervals. - Set up alerts for when your sales go above or below forecast by a certain percentage. - Share the dashboard with your team. But don’t bombard everyone with daily emails—weekly or monthly is usually enough.
What to ignore:
All the “insight” cards Empler tries to auto-generate. They’re often just surface-level stats. Focus on your core metrics.
Pro tip:
A dashboard that’s too busy gets ignored. Keep it simple: actuals, forecast, and a clear “Are we on track?” signal.
8. Iterate and Update Regularly
Forecasting isn’t a set-it-and-forget-it job. Your sales data changes, so should your forecasts.
- Update your data monthly (or more often if you have lots of sales).
- If a new product line launches or the market shifts, revisit your model.
- Don’t be afraid to switch models if your current one stops working. There’s no “forever” setting here.
Pro tip:
Keep a changelog: what you changed, when, and why. It’ll save you headaches when you wonder why the forecast started acting weird.
Keep It Simple and Keep Going
Sales forecasting with Empler isn’t magic, and it’s never perfect—but it’s better than guessing. Start with clean data, pick a model that fits your business, and don’t get lost in endless tweaking. The simpler your setup, the easier it is to spot problems and improve over time. Forecasts get better the more you use them—so get started, keep it real, and don’t let “analysis paralysis” slow you down.