Enterprise & multi-language modelling

Turnella plans large, multi-site, multi-language, multi-channel operations. This page explains how to model them honestly, including what the product does and does not do.

How do I model five languages (EN, FR, DE, ES, AR)?

Turnella has no separate "language" field. You model each language as its own workstream, then group a language's workstreams into one language pod (a Team). So for five languages on one channel you create five workstreams (one per language) and name them clearly, e.g. "EN Phone", "FR Phone", "DE Phone", "ES Phone", "AR Phone".

For a full front office (phone, email, chat, social) in five languages that is 5 × 4 = 20 workstreams per site. Create one Team per language and add that language's workstreams to it. The Enterprise report (left sidebar) then rolls every workstream up by site, channel and language into one total.

Can one language team handle phone, email, chat, social and backoffice?

Yes. That is the recommended structure — a language pod. Create one workstream per channel for that language, add them all to one Team, and set the Team to Blended so the pod is planned as one multi-channel operation. Each channel keeps its own correct staffing model (Erlang C for phone, concurrency for chat, backlog flow for email/social/backoffice).

Can I blend phone and email work?

Yes. Phone (voice) and email/social/back-office (async) blend well: the async work is absorbed into the idle time voice agents already have between calls, so you need fewer agents than staffing each separately. Set the Team to Blended to see the saving.

Can I blend chat into phone idle time?

No — keep chat dedicated. Chat is synchronous, concurrent work with its own first-response SLA. If you interleave chat with live calls the chat SLA breaks. Turnella deliberately does not absorb chat into voice idle time; chat agents stay dedicated in both the blended and separate figures. A small chat queue (one agent) can still miss its first-response SLA — add a second chat agent at peak or consolidate chat across languages in quiet hours.

How do I model social media?

Social media is modelled on the Messaging channel and labelled "Messaging / Social". It uses the asynchronous backlog/throughput engine (like email), not a real-time voice queue. Create a workstream, choose the Messaging channel, and set the average handling time and SLA for social.

How do I model second-level / escalation support?

Second-level support is modelled on the Backoffice channel and labelled "Backoffice / Second level". It uses the backlog/throughput engine with your second-level handling time (often longer than first-level) and SLA (often 24–48h). Note that Turnella treats it as its own queue — it does not automatically link second-level demand to your first-level resolution rate.

How do I model team leads who are budget-only?

Team leads who cost money but are not scheduled into the queue are added as a budget-only line, not as agent capacity. Open the Enterprise report (left sidebar) and use the "Budget-only team leads" box: enter the number of teams and the team-lead hourly rate. Turnella adds their cost (number of teams × 8h/day × operating days × rate) to the grand total and clearly marks it as not counted as agent coverage. Team leads never inflate your required FTE or schedule.

Why is my cost page showing 0 (or blank)?

Costs show as zero / blank (in your workstream's currency) because no base hourly rate is set yet — that is expected, not a bug. Open the Cost tab (or Settings → Cost model), enter a Default hourly rate and your on-cost uplift %, and every cost figure populates instantly. You only need a rate on a workstream for its cost to appear.

What happens if my vendor overhead is €25,000 per hour?

Turnella flags it. €25,000 per hour annualises to about €219,000,000 a year — many times an entire agent payroll — which is almost always a unit error. In the Enterprise report's overhead box, choose the right unit (per hour / day / week / month / year). If an hourly figure annualises to more than the whole agent payroll, Turnella shows a warning ("almost certainly a unit error — did you mean €25,000/MONTH?") and excludes it from the total until corrected. €25,000 per month (€300,000/yr) is plausible and is included.

Does Turnella support skill-based routing?

No. Turnella does not perform skill-based routing or dynamic language pooling. Each workstream is calculated as an independent queue with its own coverage floor. This is conservative: if your real operation pools multilingual agents across languages, Turnella may overstate required staffing versus a true skill-routing model. It will never silently understate it.

What is the difference between split languages and pooled languages?

• Split languages (what Turnella calculates): each language is its own queue, staffed separately, each with a minimum coverage floor. This is conservative and is the only model Turnella can actually schedule.
• Pooled languages (theoretical): one multilingual queue where any agent takes any language. Pooling is more efficient (fewer agents), but it is only achievable if your agents and routing genuinely handle all pooled languages in real time. Any "pooled" figure is a theoretical benchmark, not an achievable schedule unless you truly have multilingual routing. Splitting five languages into five queues typically needs noticeably more voice agents than one pooled queue would — that gap is the cost of independent queues.

Can I create an example weekly schedule for 5 agents?

Yes — in one click. Open a Team (language pod) and click "Generate agent schedule" in the Agent workload schedule panel. Turnella draws a visual weekly grid: rows are agents, columns are days, and each day shows colour-coded channel blocks (Phone, Chat, Email, Social/Messaging, Backoffice/Second level, Break, Off) with a legend. It shows within-day channel variety — phone at peak, email/social in the troughs, a protected back-office block, chat kept dedicated in its own colour, realistic breaks, and days off staggered across the week.

If your roster has named agents on the team's workstreams, it uses those real agents; if the roster is empty it generates clearly-labelled example agents so you can still preview the pattern. You can Regenerate, Copy, Print / PDF, and Export (CSV) the schedule. It is an advisory workload allocation — it illustrates how a blended day splits across channels and is not saved as your official shift records yet. You can also reach a single-channel example from any workstream's Schedule tab ("Generate 5-agent example schedule").

Can agents keep the same start time every day?

Yes — tick "Same start each day" on the agent workload schedule. Each agent then keeps one shift window across all their working days (the same start time between days off), which is more predictable for staff, while start windows stay staggered across agents so the team still covers the day. Turnella shows a note that this flexes less to demand, so expect some coverage thinning at the window edges (early mornings / late evenings). Leave it off for the default rotating pattern, which adapts better to peaks and troughs.

How do I save the agent schedule as a PDF?

Click Print / PDF on the agent workload schedule. It opens your browser's print dialog — choose "Save as PDF" as the destination. Turnella prints just the schedule grid (with its legend, agents and days), not the rest of the page.

How do I mark a site as internal or outsourced (vendor)?

Open Teams → the Sites panel, and use the Internal / Outsourced toggle on each site (sites default to Internal). The Enterprise report then splits outsourced vs internal staffing and cost directly from this flag — no name-guessing. You can also set optional default rates per site (agent €/h, team-lead €/h, currency, on-cost %) that new workstreams created in that site inherit.

How do I assign a workstream to a site, and why don't I see my new site?

A new site has no workstreams yet, so it shows in the left sidebar with an "add one →" link. Assign a workstream to a site three ways: when you create it (the New-workstream wizard shows a Site picker when you have more than one site); by dragging a workstream in the left sidebar onto another site's group (when you have more than one site); or any time from the workstream's Settings tab (the Site dropdown). Every site you create is listed in Teams → Sites.

How do I model annual / holiday seasonality?

Use a monthly multiplier on top of your baseline. A holiday/travel pattern is roughly: winter low ~65%, shoulder months ~90–120%, summer peak ~160% of a neutral month. Turnella's statistical forecaster learns day-of-week, week-of-month and intraday shape from history; the month-of-year curve is applied as a manual multiplier, not learned. Document your seasonal multipliers so the summer-vs-winter swing in your plan is transparent.

Is the labour cost a fully-loaded finance figure?

No — it is a planning floor. Unless you configure them, it excludes employer on-costs (social charges, pension, holiday pay), recruitment and training, facilities/IT/licences, the cost of imperfect seasonal flexing, and management above team-lead level. A fully-loaded finance figure is typically 25–40% higher. Set the on-cost uplift % on each workstream's Cost tab to include employer costs.

Related reading: Multi-language staffing · Multi-site planning