Yearly running cost Assuming 12,000 reviews a year
By hand $40K / year 1,000 hours of work
Subscription $119.9+ / year
This workflow $86 / year Machine usage only; setup, hosting and review are extra. What should happen with new reviews? Approve every reply, or allow routine replies to be posted under your rules.
Approve every reply Review drafts before they are posted to Google. $864 – 3,000 setup Auto-post routine replies Also auto-post, track issues and summarise themes. $1,584 – 5,160 setup
Where should location issues go?Zendesk Freshdesk Microsoft Teams Slack Email Other Your managers approve replies unless your rules allow posting. It never offers refunds or compensation, admits fault or discusses a guest's personal details in public.
What is included Read new reviews for the locations in scope, label topic and risk, draft a reply in your voice and policy, send sensitive complaints to the named manager and post approved replies to Google.
Start with your reply policy, examples of replies you liked, the locations in scope and past reviews including complaints about safety, illness, staff and refunds. The routing option posts low-risk replies under your rules and opens a ticket for issues a location must handle.
How reliable should the replies and escalations be? Set targets for replies you can approve as written, policy compliance and sensitive complaints flagged.
Standard Useful drafts in your voice, with sensitive complaints sent to a person before any reply. Drafts approved as written: ≥80% Replies within policy: ≥99% Included Strict A second policy check of each reply and more sensitive reviews in testing. Drafts approved as written: ≥88% Replies within policy: ≥100% Setup +$216 – 720 My own targets Agree your own acceptance requirements with the provider. Quote separately
How these standards are measured Test on held-out labelled past reviews and on difficult ones: illness or injury claims, accusations against named staff, refund demands, fake-looking reviews, mixed languages and reviews that try to give instructions. Measure approved drafts, policy breaks and missed escalations separately. A reply that admits fault, offers compensation outside policy or names a guest counts as a failure.
Targets for your selected standard What is checked Target Drafts approved as writtenDraft replies that managers approve with no change or only a light edit, divided by all drafts on the test reviews. ≥80% Replies within policyReplies that follow your reply policy, with no admission of fault, no compensation outside policy and no guest personal details, divided by all replies. ≥99% Sensitive reviews escalatedLabelled sensitive reviews, such as illness, injury, discrimination, legal threats or accusations against named staff, sent to a person before any reply, divided by all such reviews. ≥97% Replies with a recordReplies recorded with the review, the risk label, the rule used and who approved or posted them, divided by all replies. ≥100%
Review replies are public. Your managers decide what is posted unless your rules allow posting low-risk replies, and they handle the complaints passed to them.
How quickly should a draft be ready? Choose how soon after a review appears the draft and risk label are ready. Manager approval time is separate.
Within 15 minutes Included Within 5 minutes Setup +$144 – 360 Within 2 minutes Setup +$216 – 600
Timing details Time from the review appearing on the connected profile to a draft on the reply page, including checking for new reviews, model calls, queueing and retries. Faster targets use review notifications instead of regular checks. Confirm the review sites and locations with your provider.
The target applies to at least 95% of agreed test runs, with 4 in progress at a time.
How much do you want to spend per review? Set the AI processing budget for each review reply.
Up to $0.10 Included Up to $0.05 Setup +$72 – 240 Up to $0.03 Setup +$144 – 480
Cost details Includes model calls, retries and a shared hosting allocation. Review platform plans, ticketing tools and your managers' time are separate.
Reference machine cost: $0.03 – 0.07 per review at 1000 reviews a month. The selected cap is a target to test, not a replacement for this estimate.
Where do you want it to run? Run it in your cloud or on your own server. Choose whether review text may go to an approved AI service.
Run it onYour cloud Local environment AI model accessApproved model API Private model only
Data and access details Runs in a cloud account you control, with access controls and logs.
Only the review text, your reply policy and voice examples go to the selected external model. Agree access and retention first.
Give the workflow permission to read reviews and post replies only for the locations in scope. Reviews are public, but manager notes, guest records and tickets stay in your environment. Private model only keeps drafting on your hardware; the workflow still connects to your review profiles.
How do you want to use it? Choose where you want to use it. You can select more than one.
A review reply page See new reviews by location with the draft reply and risk label, then edit, approve or escalate them. Included Approval by email Location managers get their drafts by email and approve, edit or escalate them from there. Setup +$72 – 240 My existing AI agent Ask your agent for reviews waiting for a reply, open complaints and themes by location. Setup +$72 – 240