Add Train-the-Trainer To A Project

Convert a workshop you deliver yourself into a program other facilitators can run on their own. The worked example throughout is Build-a-Bot, and the structure generalizes to other Studio offerings.

Readiness criteria

A train-the-trainer program is worth considering when three conditions are met:

• the workshop already works in your hands and gets repeated requests;
• demand exceeds the days you can deliver yourself;
• and the value of the experience lives in its structure rather than in your personal facilitation.

Build-a-Bot crossed those thresholds in early 2025 after about a dozen direct deliveries.

Hold off if the workshop still mutates every time you run it, or if it only works because of expertise you cannot yet package. Stabilize the workshop in your own hands first; otherwise the handoff scales your rough edges along with the parts that work. If the expertise is hard to transfer, self-serve materials or co-delivery models are better fits than a TTT pipeline.

Three tiers of reach

A Studio program usually sits on one of three rungs. Train-the-trainer is the middle rung, and the rungs feed one another.

Tier	Description	Your involvement
1. Direct delivery	You facilitate the workshop yourself	High; you are in every room
2. Train-the-trainer	You train facilitators who then run the workshop independently	Medium initially; declines as facilitators stabilize
3. Self-serve toolkit	Anyone can pick up the materials and run the workshop	Minimal; you maintain the materials

A pipeline for potential facilitators comes from doing the direct delivery (potential facilitators can also come from the self-serve toolkit, but those people require a little more onboarding because they haven’t experienced the training/workshop directly) In Build-a-Bot, most of the facilitator pipeline came from people who first experienced the workshop as participants at ASU+GSV and SXSWedu, then asked to run it themselves (or responded that way from a post-workshop survey). Every facilitator you train stress-tests your written materials and surfaces the gaps that you can use to improve the self-serve tier.

Five phases

Phase 1: A teachable backbone

It helps to have a clear, easy-to-follow structure another person can learn and run. For Build-a-Bot the backbone is a three-move arc, Deconstruct → Design → Experiment, in which participants take apart an example bot, design their own, and run rapid test cycles. The arc fits on a slide, which matters when you are handing it off, and it gives a new facilitator clear and uncomplicated direction and helps them recover gracefully if a session starts to get away from them.

In addition to the arc, two other things should be figured out in this phase: the concrete artifact a participant leaves with (Build-a-Bot ships participants out with a working bot in bot101.app), and the timing envelope (Build-a-Bot runs cleanly between one and three hours depending on the audience). If those three pieces are not stable, Phase 2 will conceal the instability inside the facilitator kit.

Phase 2: The facilitator kit

The Build-a-Bot kit is good reference template for what a Studio TTT kit should contain:

• A facilitator script with minute-by-minute timing, written so a first-timer knows the rhythm without having seen you run it.
• Participant materials: worksheets, the R\&D Lab Book, and printable Experiment Cards.
• The platform or tool itself, kept simple. Build-a-Bot uses bot101.app, which is no-code and keeps the support burden low across a distributed facilitator network.
• A self-paced refresh: a recorded voiceover walkthrough at workshop.bot101.app that facilitators rewatch before each session they run. This is the highest-leverage component of the kit for maintaining fidelity at low cost.
• A single canonical resource folder in Google Drive holding everything above.

The kit is considered solid when a facilitator who attended the workshop once can run it from the materials alone.

Phase 3: Facilitator pipeline

Track every facilitator through four stages. A single Google Sheets spreadsheet is sufficient as the CRM;

Stage	Meaning
Lead	Expressed interest, not yet trained
Trained	Completed a facilitator kick-off
Active	Has run at least one session
Stale	No activity in 6+ months, needs re-engagement

Leads come from three sources that have produced active facilitators for Build-a-Bot:

1. post-workshop interest forms (“would you want to run this yourself?”),
2. conference attendees,
3. and referrals from partner organizations.

Capture name, organization, target audience, location, and intended timeline. The timeline field is the one that tells you who is real; a lead without a date in mind seldom converts.

Phase 4: Suggested Training cadence

The handoff happens in stages rather than all at once. Build-a-Bot and Flash Lab both use a five-step cadence that takes a new facilitator from observation to independence:

1. A training run-through (in person or over zoom) with the facilitator-in-training. (can be 1:1 or 1:many)
2. You run a dedicated train-the-trainer session in which facilitators experience the workshop as learners, and you name the moves as you make them. (It also works if they’ve just attended as an attendee)
3. They co-facilitate part of a live session with you in the room.
4. They run it solo while you observe and take notes.
5. They run it independently.
6. You reconnect after their first independent session to capture what worked and address the gaps.

A group of peers learning together generates a small community that can absorb some of the support load you would otherwise carry one-on-one.

Phase 5: Sustaining the network

Most facilitators who go stale do so in the months after training, not at training itself. Phase 5 is the part of the program most prone to being skipped, and the part most predictive of long-term reach.

Some mechanics that could help with network sustainability:

• A short check-in after a facilitator’s first independent run, which is the cheapest touch with the highest information return.
• A monthly office-hours block. Answering the same question once a month for a group beats answering it ten times in DMs.
• Newsletter and social showcases. Courtney Garza, for example, moved from trained to delivering Build-a-Bot at TCEA on her own and then posted about it.
• Story collection. Most facilitators say yes when asked to be featured, and their accounts of running the workshop are the strongest recruiting and impact material the program has.

Quality control

We are still developing how to manage quality control, but essentially, a mature quality control system for a TTT program has four layers:

1. A written standard. A definition of what running the workshop well looks like, in observable behaviors:

   • keeps the session on track when it starts to get away from them
   • models the design moves rather than describing them
   • sends participants out with the named artifact, not a partial one

   Without that document there is no shared definition of quality and no anchor for a feedback conversation.

2. A competence check before independence. Phase 4 of the cadence taken seriously: a facilitator should not run a session independently until you or a senior facilitator have watched them run it well at least once and signed off against the written standard. The current Phase 4 implies this but does not gate on it.

3. Ongoing signal collection from three triangulating sources:

   • A short post-workshop participant survey on three or four specific quality dimensions, aggregated per facilitator so trends become visible across a cohort.
   • The participant artifacts themselves. Build-a-Bot has the property that the bots participants build are inspectable evidence of what the workshop produced.
   • A recorded session every six months or so, reviewed asynchronously by a senior facilitator against the same written standard.

4. A periodic refresh. Once a year, active facilitators sit through an updated walkthrough, surface the parts of their delivery that have drifted, and recommit to the current standard. The refresh doubles as the change-management channel for material updates.

Two structural pieces support all of the above:

• A lead-facilitator tier. A small number of senior facilitators authorized to run train-the-trainer sessions, observe new facilitators, and review recorded sessions. This is how the program scales beyond your own attention without losing fidelity.
• A documented remediation path, in both directions. Facilitators whose signal drops get an explicit refresher conversation and a co-facilitation reset, not a quiet drop to stale. Facilitators who excel get a visible path into the lead-facilitator tier.

The “certified facilitator” question fits inside this architecture. Certification becomes a credible external signal once the standard, competence check, and refresh exist behind it.

Metrics

Reach numbers in isolation hide one-and-done facilitators. The metrics worth tracking together:

• Facilitators trained who became active (the conversion that matters)
• Sessions run by other people (not by you)
• Repeat facilitators
• Participants who take a defined next step after the workshop
• Quotes and stories collected

Common pitfalls

• Trying to run a TTT model on a workshop that is still being refined with large changes occurring after each delivery.
• Training facilitators and then losing contact.
• Adding tool features faster than the network can absorb the support burden.
• Capturing leads without a target timeline; those leads seldom convert.

Applying this to a new program

These principles were developed to suggest how the CRISPRkit Ambassador program could be developed. The pattern transfers as follows:

1. Identify CRISPRkit’s three-to-five-move teachable arc and the concrete artifact a participant leaves with.
2. Draft the facilitator script with timing, using the Build-a-Bot script as a structural template.
3. Record a self-paced refresh walkthrough as soon as the script is stable.
4. Stand up a single kit folder containing the script, participant materials, the tool itself, and the refresh video.
5. Open an interest form and start the Lead → Trained → Active → Stale tracker.
6. Run the first cohort kick-off, then proceed through co-facilitation, observation, release, and debrief.
7. Schedule the sustain rhythm before the first cohort goes independent: first-run check-ins, monthly office hours, story collection.

The Ambassador framing is just a relabeling. “Ambassadors” are “facilitators,” and the playbook above could potentially transfer with little structural change once Phase 1 is developed for CRISPRkit’s content.