The handover is the part most vendors get wrong
For the first two to four weeks after cutover, hypercare, the team that built and migrated your program stays on it. A dedicated support team then takes it over, monitoring the platform 24/7 and handling first, second and third line requests. Second and third line work routes to core engineering, and the implementation specialists are brought back when a question needs them; both are internal assignments, so you raise one ticket and never chase it between teams. Every program has a named account manager and a regular business review. Managed services are available where a client wants day-to-day program operations run for them.
Hypercare, then a planned handover
Most vendors change the people at cutover and call it business as usual. The program goes to a support organisation that was not in the room for any of the decisions, and the client spends the next quarter explaining their own program back to their vendor.
Hypercare
The team that built and migrated your program stays on it through cutover and the first weeks of live running, while the first real cycle of member activity, promotions and settlement passes through the platform.
A dedicated support team takes it over
A team assigned to your program picks it up, working around the clock. They monitor the platform and handle first, second and third line requests.
The context, not just the queue
Handover is a piece of work with its own exit criteria, not an email. The configuration decisions, the systems map from Vision Alignment and every open item move with the program, so the team answering you has read it before you write.
How the program gets to this point is described on Implementation and Ownership.
What is watched
Loyalty is a real-time financial system, and it is watched 24/7. Most of what matters is not whether the platform is up, but whether the program is behaving the way it was configured to behave.
Platform health
Availability, API response time, error rates and regional failover state, watched against the service levels agreed in your contract. Targets differ by engagement, so the SLAs that apply to your program are the ones monitored.
Program behaviour
Rule execution volumes, promotion evaluation outcomes, tier movement, expiry runs and accrual patterns. A promotion that stops firing is a program incident even when every system reports healthy.
The edges
Event delivery to downstream systems, partner feed processing through FeedXChange™, batch job completion and integration endpoints. Most production surprises arrive through an edge rather than the core.
Monitoring covers the platform Loyalty Methods operates. Systems you own, and the channels that call into ReactorCX, stay on your side of the boundary; the architecture review walks through exactly where that line sits for your estate.
How you reach us
Three channels, one queue, covered 24/7. The channel you choose changes how fast you can explain the problem, not who sees it or how it is ranked.
Chat
For questions with a short answer and for anything easier to describe by pasting a configuration or an identifier.
Phone
For anything affecting members in production, and for the point in an incident where writing it down is slower than saying it.
For requests that carry attachments, need a written trail, or are not urgent enough to interrupt anyone.
Severity sets the path
- Severity one — members are affected in production: balances wrong, redemptions failing, enrolment down. Incident response opens immediately and the account manager is notified without being asked.
- Severity two — the program is degraded but members are not visibly affected: a promotion misfiring, a feed backing up, a report not refreshing.
- Severity three — something is wrong but contained, or a change is needed that is not urgent.
- Consultative — nothing is broken. See below.
Cover does not stop outside office hours. Someone is on duty at all times, and what happens next is decided by the severity of the issue rather than the hour it arrives.
How a request reaches the right person
You raise one ticket. Getting it to the right desk is our routing problem, not yours.
Every step of that routing happens inside Loyalty Methods. You are not asked to raise the request again with a different team, you are not handed a second reference number, and nobody asks you to explain the program from the start because the ticket changed desks.
Two kinds of ticket
Most support organisations are built to answer one question: is the software behaving as specified? That question has a yes or no answer, and once the answer is yes the ticket closes.
Loyalty teams spend far more time on a different question. How should we model a partner-funded offer so the settlement lands in the right place? What happens to tier qualification if we change the earn rate mid-quarter? Is there a cleaner way to stage this promotion than the one we have built?
Those are consultative tickets. They go to the same team, through the same queue, and are answered by people who can open your configuration and look. They are not redirected to an account executive, and they do not require a statement of work to get an answer.
Break-fix
Something is not working as configured. Resolution restores expected behaviour, and the execution log shows what ran and why.
Consultative
Everything is working; the question is how best to use it. Resolution is a recommendation, usually with the configuration drafted alongside it.
Who you work with
A named account manager
One person who owns the commercial relationship, chairs the business review and is the escalation route when something needs to move faster than a ticket. Named, not a rota.
A defined support team
A set of engineers assigned to your program rather than a general pool, so context does not have to be rebuilt on every ticket.
The implementation specialists
The people who built your configuration are brought back when a question needs them, as an internal assignment rather than a new engagement.
Managed services
ReactorCX is built so your team can operate it without engineering help. Some clients would still rather not staff that work, particularly in the first year or through a period of heavy campaign activity.
Where that is the case, Loyalty Methods can run the day-to-day: building and scheduling promotions, managing configuration releases through stage into production, producing program reporting and watching program outcomes rather than only platform health.
It is an option and it is reversible. Programs move onto managed services and off them again, and the platform does not behave differently either way.
What stays yours
- The program strategy and the calendar
- Approval of every change before it reaches production
- The data, the ledger and the audit trail
- The ability to take the work back in-house
A standing conversation about the program, not the platform
Platform health belongs in monitoring. The review exists for the questions monitoring cannot answer.
What gets reviewed
- Program performance against the objectives set in Vision Alignment
- Liability position and how it is moving
- Open issues, recurring tickets and anything that keeps coming back
- Platform releases since the last review and what they make possible
What comes out of it
- An agreed list of changes with owners on both sides
- Roadmap input that reaches the people who build the platform
- Early warning on anything that needs budget or a project
Frequently asked questions
What happens after a loyalty program goes live on ReactorCX?
For the first two to four weeks, hypercare, the team that built and migrated the program stays on it. A dedicated support team then takes over, monitoring the platform 24/7 and handling first, second and third line requests, with core engineering and the original implementation specialists brought in internally when a question needs them. Each program has a named account manager and a regular business review.
What is hypercare, and how long does it last?
Hypercare is the period straight after cutover when the implementation and migration team stays on the program rather than handing it over immediately. It runs for two to four weeks, covering the first real cycle of member activity, promotions and settlement. Handover to the dedicated support team happens at the end of it, against its own exit criteria rather than a date alone.
How do I raise a support ticket?
Through chat, phone or email. All three open a ticket in the same queue, so the channel you choose does not change who sees the request or how it is prioritised. Each ticket is assigned a severity, and severity sets the response path.
Can I ask how to model something, rather than report a fault?
Yes. Those are handled as consultative tickets and go to the same team. A question about how to structure a tier, model a partner-funded offer or stage a promotion is a support request, not a sales conversation, and it is answered by people who know the configuration.
Does Loyalty Methods offer managed services?
Optionally. Where a client would rather not staff day-to-day program operations, Loyalty Methods can build promotions, manage releases and run reporting against the program. It is an option, not a dependency: the platform is designed for client teams to operate it themselves.
Is support available outside business hours?
Yes. Cover is 24/7 and someone is on duty at all times. What happens next is set by the severity of the issue rather than the hour it is raised: a problem affecting members in production opens incident response whenever it arrives.
Who is my point of contact?
A named account manager, supported by a defined support team rather than a general pool. The account manager owns the commercial relationship, the review cycle and escalation, and is the person to call when something needs to move faster than a ticket.
Ask the people who would be supporting you
A support conversation before you buy is a better test of a vendor than a reference call. Bring the awkward scenario.