E-commerce
July 1, 2026
"The bot recommends a trial kit outside of the sensitive skin range." "The AI promises a refund without mentioning the return_window." "Trial finished, chatbot does not offer the full size." Three failures where a poorly calibrated guided selling trial kit bot ignores TRIALKIT-MAP or confuses the starter kit.
An AI trial kit chatbot does not replace TRIALKIT agents (#537). It reads TRIALKIT-MAP, recommends a matching profile kit, mentions the return window, and guides the transition to full size.
This guide #538 covers bot_trial_kit_* intents, the TRIALKITbot flow, and trialkit_bot KPIs. Pair bot from the TRIALKIT playbook (#537). New AI use case: guide to the right kit then to the right product post-trial.
Summary
Why automate the trial kit with a bot?
"Which trial kit for a beginner?" arrives in PDP pre-purchase. A calibrated bot reads TRIALKIT-MAP, matches profile, and explains TBYB return_window without promising refunds outside of policy.
What the tier 1 bot resolves
Kit recommend: trial_kit_sku skin hair profile tags
Content explain: trial_duration ritual_steps_en
Return window: return_window_days return_deadline cite
Convert path: full_size_upgrade_sku PDP link
Post-trial branch: keep vs return vs convert
Try-before-buy apps expose trial status and return deadline (Shopify TBYB 2026). The bot intercepts bot_trial_kit_recommend before guided selling #150 alone.
TRIALKITbot vs TRIALKIT #537, SAMPbot #536, starter kit and guided #150
Five contents, five product trial layers.
Quick matrix
#538 TRIALKITbot: bot recommend kit convert tier 1
#537 TRIALKIT: agents return refund reaction
#536 SAMPbot: free mini conditions
Starter kit: paid pack credit upgrade
Guided selling #150: full size SKU pre-purchase
Router: trial kit TBYB → #538. Complex return refund → #537 agents. Free sample → #536 SAMPbot.
Which bot_trial_kit_* intents should be configured?
Eight intents bot trial kit.
Eight intents bot_trial_kit
bot_trial_kit_recommend: which trial kit for my profile
bot_trial_kit_contents: what does the duration kit contain
bot_trial_kit_return: how to return deadline
bot_trial_kit_convert: buy full size post trial
bot_trial_kit_keep: keep kit payment capture
bot_trial_kit_ritual: directions for use instructions ritual
bot_trial_kit_deadline: return deadline limit keep
bot_trial_kit_escalate: reaction refund handoff #537
Tier 1 auto: recommend, contents, return, convert, ritual, deadline if TRIALKIT-MAP lookup.
Intent bot_trial_kit_escalate → agents #537 with payload trial_kit_reaction order_id.
How should I consume TRIALKIT-MAP #537?
The bot reads TRIALKIT-MAP #537: trial_kit_sku, trial_duration_days, return_window_days, keep_or_return_policy, full_size_upgrade_sku, partial_open_return, conversion_credit_eur, ritual_steps_fr.
Lookup grounded
Profile match: kit skin hair tags vs quiz answers
Order lookup: trial_status delivered_at return_deadline
Pre-purchase gate: recommend in-stock kit SKU only
Convert branch: full_size_upgrade_sku credit cite
Return branch: return_window_days portal URL
Alignment anti-hallucination (#123): never kit off TRIALKIT-MAP whitelist.
TRIALKITBOT-SUP policy in six rules
Six rules bot kit safe trial.
TRIALKIT-MAP-GROUNDED: recommend from map kit_sku tags
RETURN-WINDOW-CITE: return_window_days exact deadline
NO-REFUND-PROMISE: refund cite keep_or_return_policy
CONVERT-PATH-CITE: full_size_upgrade_sku PDP mandatory
STARTER-DISTINCT: starter kit u2192 reroute starter playbook
ESCALATE-REACTION: trial_kit_reaction u2192 #537 agents
Flow TRIALKITbot TKB-1 to TKB-8
Flow eight steps bot trial kit.
TKB-1 Classify: bot_trial_kit_* intent
TKB-2 Context: pre post purchase profile order
TKB-3 TRIALKIT-MAP: kit return convert ritual
TKB-4 Recommend: profile match top kit in stock
TKB-5 Guardrail: RETURN-WINDOW NO-REFUND-PROMISE
TKB-6 Respond: TPL-TRIALKITbot grounded
TKB-7 Convert CTA: full_size link if post-trial
TKB-8 Log: kit_offered convert_tag escalate
Example TPL-TRIALKITbot-RECOMMEND
"Profile [profile]: kit [trial_kit_name]. [trial_duration_days] days of use. Return within [return_window_days] days. Full size after trial: [full_size_upgrade_sku]. PDP: [pdp_url]."
TPL-TRIALKITbot and touchpoints templates
Four essential templates.
TPL-TRIALKITbot-RECOMMEND
Profile [profile]: [trial_kit_name]. Content [items_summary]. Ritual: [ritual_steps_fr].
TPL-TRIALKITbot-RETURN
Return before [return_deadline]. Portal: [return_portal_url]. Refund according to [keep_or_return_policy].
TPL-TRIALKITbot-CONVERT
Successful trial? [full_size_upgrade_sku]: [pdp_url]. Credit [conversion_credit_eur] € if applicable.
TPL-TRIALKITbot-DEADLINE
D-[days_left]: return or keep kit. Keep = capture [capture_info]. Convert: [pdp_url].
Touchpoints
PDP trial kit chip: bot_trial_kit_recommend quiz
Post-delivery D+1: bot_trial_kit_ritual proactive
D-3 deadline: bot_trial_kit_deadline convert CTA
Post-trial D+7: bot_trial_kit_convert full size
Edge cases and reroutes
Five cases outside tier 1 bot.
Free sample requested: reroute #536 SAMPbot
Starter kit confused: STARTER-DISTINCT starter playbook
Skin reaction: ESCALATE-REACTION #537
Refund execution: handoff #537 finance agents
Full size pre-purchase: reroute guided selling #150
Bot does not capture TBYB payment directly: explain keep_or_return_policy, handoff #537 if dispute.
Essential trialkit_bot KPIs
Five TRIALKITbot management metrics.
trialkit_bot_recommend_rate: % recommend compliant with TRIALKIT-MAP
trialkit_bot_add_to_cart: kit added post-recommend
trialkit_bot_convert_click: full_size_upgrade clicks post-trial
trialkit_bot_window_compliance: RETURN-WINDOW-CITE 100%
trialkit_bot_escalate_reaction: sessions → #537
Monthly red team: 10 prompts "guaranteed refund without return", "kit outside description for sensitive skin". Refund out of policy = NO-REFUND-PROMISE violation.
Anti-patterns TRIALKITbot
Six common mistakes.
Kit not on whitelist: TRIALKIT-MAP-GROUNDED violation
Refund promised tier 1: NO-REFUND-PROMISE
Deadline not cited: RETURN-WINDOW-CITE
No convert path: CONVERT-PATH-CITE mandatory
Confusing SAMPbot #536: distinct free sample
Confusing starter kit: STARTER-DISTINCT
TRIALKITbot with Qstomy
Qstomy on Shopify: detect bot_trial_kit_* intent, TRIALKIT-MAP RAG, profile quiz match, return deadline widget, convert CTA, handoff #537 escalate.
Pipeline: bot recommend kit → #537 return refund → #150 full-size guided selling → AI governance #142 audit convert claims.
Explore AI support and request a demo.
Checklist, FAQ and going further
TRIALKITbot Checklist (8 steps)
Sync TRIALKIT-MAP #537: RAG bot kit tags profile
Policy TRIALKITBOT-SUP: 6 RETURN-WINDOW-CITE rules
8 intents bot_trial_kit_*: flow TKB-1 to TKB-8
4 templates TPL-TRIALKITbot-*: RECOMMEND RETURN CONVERT DEADLINE
PDP quiz profile match: recommend grounded
D-3 deadline webhook: bot_trial_kit_deadline proactive
Red team 10 prompts: refund outside policy forbidden
Dashboard KPI: trialkit_bot_* section 9
FAQ
Difference #537?
#537 = agents return refund reaction. #538 = bot tier 1 recommend convert guide.
Difference SAMPbot #536?
#536 = mini free. #538 = kit TBYB paid deferred.
Bot promises refund?
Cite keep_or_return_policy only. ESCALATE #537 if dispute.
Full size pre-purchase?
Reroute guided selling #150, not trial kit.
Go further
p: Starter kit
This week: index TRIALKIT-MAP in corpus bot, configure PDP quiz profile match, test D-3 deadline proactive flow, measure trialkit_bot_convert_click.
Enzo
July 1, 2026
