Draco · Skyline Blog
How to Get Your Agency's CRM Ready for December
Fabián Alvarez8 min read
TL;DR
- Start preparing four to six weeks before December, not in the final week.
- Move Vega from copilot to autonomous at night gradually, reverting if something fails.
- Use medium intensity on new bookings and low intensity on refunds, always with a message cap.
- Get approved WhatsApp templates ready before the peak, since Meta's approval takes days.
- Distribute fixed tasks with schedules and owners so no hot lead goes without follow-up.
Getting your CRM ready for December means mapping daily message schedules, gradually moving Vega from copilot to autonomous at night, adjusting intensity per funnel (medium for bookings, low for refunds), preparing approved WhatsApp templates weeks ahead, and assigning fixed, scheduled tasks to the human team.
Why does December break a travel agency's CRM?
December multiplies incoming messages and shrinks the team's reaction time. An agency that handled thirty chats a day in October can see triple that in the first week of December, with half arriving at night or in the early morning. If the CRM isn't ready for that volume, hot leads go cold while someone on the team is asleep or busy with another conversation. Draco, Skyline's AI CRM, is built to absorb that spike without burning out the human team: WhatsApp, Instagram, Messenger and TikTok arrive in one inbox, and a team of AI agents helps hold the pace when humans can't cover every hour.
This guide isn't a list of good intentions. It's a concrete sequence of steps so an agency — picture a receptive agency in San Andrés selling beach and diving packages — reaches December 1st with the CRM ready: schedules defined, Vega running on its own at night, funnel intensity adjusted, templates approved and tasks distributed across the team.
When should you start preparing the CRM for December?
Preparation doesn't happen on November 30th. The ideal window is four to six weeks earlier, because some pieces depend on external approvals (WhatsApp templates are approved by Meta and that takes days) and because the team needs time to trust that the AI is replying well before handing it control of the night.
Week 1: diagnosis
- Review how many messages arrived last December, by channel and by hour.
- Identify the hours when the human team isn't currently available.
- List the funnels you'll use in December: new bookings, confirmed-traveler questions, refunds or changes.
Weeks 2 to 3: configuration
This is where schedules, intensity and templates get adjusted. We walk through each in detail below.
Week 4: testing and training
The team approves or corrects what the AI proposes, and those corrections feed Draco before the real peak hits.
How do you map schedules by day before the peak?
Not every December day follows the same message pattern. A receptive agency usually sees more traffic on weeknights (people planning the trip after work) and earlier peaks on weekends (families organizing a long weekend). Before automating anything, it's worth mapping this day by day:
- Monday to Thursday: steady traffic from mid-morning until midnight.
- Friday: a strong peak right after work, between 5 and 9 pm.
- Saturday and Sunday: scattered traffic all day, with a bump after 10 pm.
With that map, the human team covers the highest-conversion hours (when a fast, personal reply closes more sales) and the AI covers the valley hours and the early morning, which is exactly where the most gets lost without automation.
How do you put Vega on autonomous mode at night without losing control?
Vega is the Draco agent that handles, pre-qualifies and closes in real time: it recommends from the catalog, quotes, builds packages and carries the sale through to payment. Draco doesn't force you to hand over full control on day one. It starts in copilot mode, where the AI drafts the reply and a person on the team approves or corrects it before sending, and moves to autonomous once the team trusts what it proposes; that change can be reverted at any time if something doesn't add up.
For an agency that wants to reach December with Vega running solo at night, the recommended path is:
Step 1: copilot during office hours
For two or three weeks, keep Vega in copilot on the new-bookings funnel, during the hours the team is present. Every approval or correction the team makes trains the AI for upcoming conversations.
Step 2: autonomous during valley hours
Once the proposed replies barely need correcting, move Vega to autonomous only during the hours when no one on the team is online, for example from 11 pm to 7 am.
Step 3: fully autonomous in December
If the previous step worked without surprises, by December Vega can stay autonomous all night, every night, while the human team takes back control during peak-traffic hours, where closing benefits from a more personal touch.
The change can be reverted at any time if the team spots something that isn't working as expected, without losing what's already configured in the funnels.
What intensity should you use in December?
Vega's intensity (low, medium, high) and Leo's (soft, medium, hard) are set per funnel, with a hard cap of messages per reply and a fixed stance of never chasing the customer no matter what. This matters especially in December, when lead volume rises and the risk of overwhelming someone with messages rises too.
For most agencies, the recommendation for December is to run medium intensity on the new-bookings funnel:
- It's active enough not to leave hot leads unanswered overnight.
- It isn't aggressive enough to trigger complaints in a month when people already get messages from dozens of different businesses.
- It keeps the cap on messages per reply, so no traveler gets a burst of texts even as overall volume climbs.
If the agency has a separate funnel for refunds or date changes (common in December due to last-minute cancellations), that's where it's worth lowering the intensity: these are conversations where the customer is already somewhat frustrated, and insisting doesn't help.
| Funnel | Recommended December intensity | Why |
|---|---|---|
| New bookings | Medium | Balances response speed without overwhelming the lead |
| Cold lead follow-up | Medium to high | December is the short window to re-engage them before year-end |
| Refunds and changes | Low | The customer is already in contact; no need to chase them |
How do you get WhatsApp templates ready before December?
Meta-approved WhatsApp templates let you reach out first outside the 24-hour window, that is, once the period during which you can freely reply to a customer has already passed. In December this is key: there are travelers who wrote in November asking about prices and never confirmed, and the only way to pick that conversation back up without waiting for them to write first is with an approved template.
Which templates to prepare before the month starts
- Booking confirmation: with date, meeting point and what the traveler should bring.
- Pending balance reminder: for those who left a deposit and still owe the rest before the tour date.
- Cold quote reactivation: for someone who asked in November and never wrote back.
- Operational change notice: in case a schedule needs to shift due to weather or availability, common in high season.
- Year-end greeting with a January offer: to keep the relationship alive after the December trip ends.
Since Meta's approval takes days, these templates need to be ready during the configuration phase, not during peak week. An agency that leaves this until December 5th risks not having the template approved exactly when it's needed most, right when the most November cold leads could be reactivated.
What tasks should stay with the human team?
Draco has tasks with a calendar for the human team, and the AI doesn't create tasks on leads that are already won or lost, so the task board stays focused on what actually needs a person. In December, before the peak, it's worth distributing a set of fixed tasks with the team:
- Review every morning the leads Vega handled autonomously overnight and flag the ones that need human follow-up.
- Manually confirm bookings for large or corporate groups, which usually need a call in addition to the chat.
- Oversee the refunds funnel, where the AI's lower intensity leaves more weight on the human conversation.
- Check the board every afternoon before the shift ends, so no hot lead is left without a task assigned for the next day.
Assigning these tasks with specific names and times — who checks the board at 8 am, who confirms groups at noon — keeps the whole team in December from assuming "someone else" already checked something.
Example: a receptive agency in San Andrés
The agency in this example sells beach and diving packages with a team of four. In October they started putting Vega in copilot on the new-bookings funnel, reviewing every proposed reply before sending it. By the third week of November, Vega had already moved to autonomous during early-morning hours, at medium intensity. The team got the confirmation and balance-reminder templates ready, and split the morning board review between two people who rotate weekly. When December arrived, the human team focused on afternoon and early-night hours, which is when year-end package sales convert best, while Vega held down the early morning without anyone needing to stay awake to reply.
What to do now
Log into your Draco account and decide today the exact hours you'll put Vega on autonomous mode during December's early mornings; the rest of the configuration — intensity, templates and tasks — gets adjusted around that first decision.
Frequently asked questions
How far ahead of December should I start configuring the CRM?
Four to six weeks ahead, because WhatsApp templates need Meta's approval and the team needs time to train the AI in copilot mode before handing it nighttime control.
Is it safe to leave Vega fully autonomous overnight in December?
It's a gradual process: first copilot, then autonomous only during valley hours, then fully autonomous in December if there were no surprises. The change can be reverted at any time.
Why use medium intensity instead of high in December?
December overwhelms travelers with messages from many businesses at once. Medium intensity responds fast without risking chasing the customer, and the cap on messages per reply protects that stance.
What happens if I don't prepare WhatsApp templates in time?
If a template isn't approved when you need it most, you won't be able to re-engage cold leads outside the 24-hour window right at December's peak, which is exactly when the most conversion is lost by not resuming those conversations.
Does the AI take tasks away from the human team in December?
No. The AI doesn't create tasks on leads already won or lost, so the task board stays focused on what truly needs human attention, like large groups or refunds.
Draco
See Draco in action
The AI CRM for agencies and tour operators: WhatsApp, Instagram, Messenger and TikTok in one inbox, handled by five AI agents.
Try Draco
Related · Draco
Comment-to-DM: Turn IG and TikTok Comments into Sales
Comment-to-DM sends an automatic direct message to anyone who comments on an Instagram or TikTok post.
ReadHow to Tune a Sales AI's Intensity Without Chasing Customers
Vega (low, medium, high) controls how proactive she is in real time within an active conversation.
ReadFunnel, Stage, Temperature, Task: Tourism CRM Vocabulary
A funnel is the sequence of stages a lead goes through, with its own handling rules for each type of sale.
Read