How to Forecast Support Ticket Volume with Zendesk Data

If you’ve ever been blindsided by a Monday morning ticket avalanche or scrambled to cover a product launch, you already know why forecasting matters. Predicting ticket volume isn’t about building a perfect model — it’s about having a good-enough estimate to staff correctly, set expectations, and avoid reactive firefighting.

This post covers a practical approach to ticket volume forecasting that works for small and mid-size support teams using Zendesk.

Why forecasting matters for support ops

Without a forecast, you’re always reacting:

Understaffed days lead to growing backlog, longer first reply time, and frustrated customers.
Overstaffed days waste payroll and leave agents underutilized.
Surprise spikes (product launches, billing cycles, incidents) catch the team flat-footed.

A basic forecast doesn’t eliminate surprises, but it turns most of them into expected events you’ve already planned for. That’s the difference between reactive ops and proactive ops.

What data you need from Zendesk

To forecast accurately, you need at least 3 months of daily ticket creation data. More is better — 6–12 months lets you spot seasonal patterns. Pull this from Zendesk Explore:

Metric: Created tickets (count)
Time dimension: Created date, by day
Optional breakdowns: Channel, group, tag, or ticket form — useful if different segments have different patterns

Export this to a spreadsheet, or use Explore’s built-in forecasting (available on Professional and Enterprise plans).

Key patterns to look for

Day-of-week pattern — Most teams see Monday as the highest volume day (customers accumulate issues over the weekend). Friday is often lower. Quantify the pattern: if Monday averages 120 tickets and Friday averages 80, that’s a 50% swing you need to staff for.
Monthly cycles — SaaS companies often see spikes around billing dates (1st, 15th) or trial expirations. E-commerce teams see patterns around promotions, paydays, and season changes.
Seasonal trends — Holiday seasons, back-to-school, year-end, tax season — these create predictable surges depending on your industry. See seasonal support planning with data for more.
Growth trend — If your customer base is growing, ticket volume trends upward over time. Separate growth from seasonality: a 10% year-over-year growth rate means next March will have more tickets than this March, even with the same seasonal pattern.

A simple forecasting method

You don’t need machine learning to forecast well. Here’s a straightforward approach:

Step 1: Calculate your baseline

Take the last 4 weeks of daily ticket counts and calculate the daily average. This is your baseline.

Step 2: Apply the day-of-week multiplier

Calculate the average for each day of the week over the last 8–12 weeks. Express each day as a ratio to the overall daily average:

Day	Avg tickets	Multiplier
Monday	115	1.28
Tuesday	100	1.11
Wednesday	95	1.06
Thursday	88	0.98
Friday	72	0.80
Saturday	40	0.44
Sunday	30	0.33

To forecast a specific day: Baseline daily average × Day-of-week multiplier = Expected tickets

Step 3: Adjust for known events

Layer in events you know about:

Product launches — Historically, how much did ticket volume increase during the last launch? Apply that multiplier.
Billing cycles — If the 1st of the month spikes 20%, add that to your estimate.
Marketing campaigns — A big promotion brings in customers, who generate tickets. Estimate based on prior campaigns.
Holidays / closures — Volume typically drops during closures but spikes the day after.

Step 4: Apply the growth trend

If ticket volume has been growing 5% month-over-month, your 30-day forecast should be ~5% higher than the same period last month (after controlling for seasonality).

Using Zendesk Explore’s built-in forecasting

If you have Zendesk Explore Professional or Enterprise, you can use the built-in forecasting feature:

Create a report with Created Tickets by day or week.
Enable the Trend Line visualization with the Holt-Winters forecasting model.
Explore uses your historical data to project future values, accounting for seasonality automatically.

The Holt-Winters model works best with at least two full seasonal cycles (e.g. two years for annual seasonality, or two months for weekly seasonality). With less data, the simple method above is often more reliable.

For larger teams, Zendesk’s Workforce Management (WFM) add-on provides more advanced forecasting with machine learning algorithms (Prophet, XGBoost, NeuralProphet) and converts forecasts into staffing schedules.

Turning forecasts into staffing decisions

A forecast is only useful if it changes what you do. Here’s how to operationalize it:

Calculate tickets per agent per day — Divide your target daily resolution volume by the number of agents available. If you resolve 90 tickets/day with 5 agents, that’s 18 tickets per agent. See tickets per agent in the glossary.
Compare forecast to capacity — If Monday’s forecast is 130 tickets and your capacity is 90, you’re 40 tickets short. That’s either overtime, schedule shifts, or accepted backlog growth.
Build a staffing model — Map forecast volume to required headcount: Required agents = Forecasted tickets ÷ Tickets per agent per day. Add a buffer (10–15%) for breaks, meetings, and variability.
Review weekly — In your weekly support ops review, compare actual volume to your forecast. Were you within 10%? 20%? Persistent under- or over-estimation means your model needs tuning.

Common forecasting pitfalls

Confusing created and solved tickets — For forecasting incoming volume, use created tickets, not solved tickets. Solved tickets reflect your capacity, not demand.
Ignoring outliers — A single incident that generated 500 tickets in a day will skew your averages. Identify and exclude (or separately model) outlier days.
Over-engineering — A spreadsheet with day-of-week multipliers and event adjustments is often better than a complex model you don’t maintain. Keep it simple until you’ve outgrown simple.
Not updating — A forecast built in January doesn’t work in July. Recalculate your baseline and multipliers monthly.

Key takeaways

Forecasting support ticket volume is about pattern recognition, not precision. Start with your historical data, identify the patterns (day-of-week, monthly cycles, growth), and layer in known events. Update regularly, compare forecast to actual in your weekly support ops review, and use the numbers to make staffing decisions before you’re underwater.

For a broader view of the metrics that feed into capacity planning, see support capacity planning and support metrics dashboard.

Track ticket volume trends without spreadsheets — start free