Agencyonboarding

Work through the stages in order. Each has prerequisites, an owner on each side, a desired output, and a risk if skipped.

The stages are sequenced by what has so far determined whether a rollout works, rather than by what’s easiest to schedule. We find agencies routinely skip Stage 2 (identity and mail) and Stage 5 (baseline) because they feel like paperwork next to a client deadline - but we have seen they are the two stages that forecast the overall outcome best.

They are strong recommendations rather than gates. If you need to waive one to meet a launch timeline, do it knowingly and make sure you revisit it as soon as possible.

Nothing here requires the full sequence to run before value appears. Stage 4 includes a path to start working before your contact data is ready.

Start with Stage 0. It is a discovery conversation, not paperwork, and it belongs in the first or second call we have with you. It exists because almost everything that delays a rollout is knowable on day one and does not get asked.

This is the methodology behind how we train users. Read it before you plan your rollout, because it changes how you group people and what you and we will ask of them.

Start with the most used end-to-end workflows

Theme one, create baseline comfort for sending messages. Every user, without exception, becomes comfortable and confident with the core announcement workflow end to end: building a list, drafting, personalising, sending, and reading what came back. This is the single most important thing to start with.

Most of the value the platform delivers comes from everyone being competent using the day-to-day workflows, not from being an expert user of infrequent ones.

Theme two, sub-groups that form around a real requirement. On top of the baseline, small groups form around what people actually need to do: watching the news cycle and reporting on it, keeping announcement flows running cleanly, or reviewing performance and measurement. Membership follows the requirement, not the org chart, and the groups do not need naming up front.

Why we don’t believe in creating power users

We used to recommend this, and have changed our position for two reasons:

• A single trained generalist is a single point of failure. When they are on leave, or they leave, the account’s capability goes with them. Several partial experts give you redundancy.
• People arrive with different use cases anyway. We’ve found that genuine interest predicts who will go deep far better than a job title does. Exploration and innovation are also always transferable: what one team member discovers will be reusable by another to achieve a different goal.

Pair every enthusiast with a sceptic

Pair whoever is moving fastest with a colleague who does not reach for AI tools first.

The value is that the sceptic forces the enthusiast to reconstruct how they got from A to Z. People moving fast with AI assistance often cannot articulate the path they took, and unarticulated reasoning is lost judgement. Remove all the friction and it removes some of the thinking with it. Adding sceptics and non-technical users in alongside AI-literate team members will result in more explainable processes and that will help other users onboard faster.

This means: your less engaged people are not a remedial cohort to throw extra training at. They’re the judgement cohort. Treat them that way and you get better output from both sides of the pair.

Account and Hub leads are the Decider

Exploring everything is a waste of time. Account and Hub leads decide which rabbit holes are worth going down, so exploration compounds instead of scattering.

What matters here is that you want to enable your team to build a deeper understanding of what you already have, rather than acquiring more tools.

What this asks of you

Whoever ends up in the performance and measurement group owns the data between the check-ins in Stage 9, so make sure that group has a name against it.

Three phases, with different audiences. They are deliberately separable: your IT Administrator does not need the enablement material, and your Account Users do not need the permission-scope detail. Share each phase with the right people, but educate them on what will occur in the other phases - your Account Leads and Users will want to know that the platform meets both GDPR and IT Security compliance standards for example.

Phase 1 begins as procurement concludes.

Within Phase 2, Account Leads own Stages 4, 5 and 7 (assets, baseline, choosing the client moment, mapping approvals). Hub Leads and Account Users own Stages 3 and 6 (accounts, users, inboxes, lists, announcements, training).

Work through this on the first or second call. You will not know every answer, and that is fine and expected. An unknown is a useful output: each one becomes an action with a name and a date against it. What causes damage is an assumption, not a gap.

Where a question looks like it is for IT, it is. Get them in the room for section A rather than relaying answers back and forth.

The five answers that change the plan most

1. The mail platform for every organisation that needs access
2. Whether your own identity policy permits generic or shared sending addresses
3. Whether your end client has a separate tool-approval process
4. Whether contacts can lawfully leave your existing media database under its licence
5. Whether pitches go out as plain text or branded HTML, and who signs the templates off

Get those five early and the rest is largely scheduling.

Who answers what

The question groups split by audience, so you can forward only the relevant part to each team.

Groups A to C are the Phase 1 pack. Send them to IT and security ahead of the call so they arrive with answers rather than taking them away.

A. Sending email: platform, protocol and identity

The section that matters most. MVPR sends and receives real email from real mailboxes, so this determines whether the core function works at all.

B. Who needs access

C. Approvals and governance

D. Contact data

E. Content, templates and how pitches look

F. Client involvement and approvals

G. Ways of working

H. Measurement

If the answer to most of section H is “not really”, that is common, and it is exactly why Stage 5 exists. Say so now rather than at the review.

Turning the answers into a plan

• Every unknown becomes an action with an owner and a date.
• The five headline answers determine the critical path. If any of them is bad news, we would rather know in week one and plan around it than hit it in launch week.
• We come back to you with a workback plan from your target go-live date, showing what has to happen when, and what we need from whom.

Realistic elapsed time for a multi-market agency rollout.

Two things run in parallel with everything else and should start at day one: compliance (Stage 1) and identity readiness (Stage 2). If they run in series after commercials, you will lose weeks.

What we need on your side

Five roles, and only five. Everything in this guidebook is owned by one of them.

Three responsibilities layered on top

These are not extra people. They are hats worn by the roles above, and they come from the methodology.

What you get from us

• A named implementation contact for the whole project
• Technical lead for permissions, data ingest and debugging
• Enablement delivery, recorded and indexed
• A shared support channel (Teams or Slack)
• Compliance and security documentation pack
• Data ingest and segmentation support
• Reviews at 30, 60 and 90 days
• Ongoing data support and measurement education

Large agency groups typically have three or four separate approval paths, and they are not aware of each other:

1. Information security review
2. AI governance or AI compliance review, often a separate questionnaire and a separate queue
3. Data protection review, which may come back with specific questions about how journalist interaction data is used
4. Your end client’s own AI council or tool-approval process, which is easy to miss until late

There may also be a procurement or vendor-onboarding track running alongside. That is usually the fastest of the four.

Notes from experience

• Ask about the end-client approval path in week one. Discovering it late can pause a rollout days before training.
• The clearest position on contact data: you are the data controller, we are the data processor. Sharing media contacts with us has a legitimate-interest basis for media relations, and our DPA covers it.

This stage is not administrative. MVPR sends and receives real email from real mailboxes, so how identity works in your organisation determines whether the core function works at all. In previous rollouts this has been the largest single cause of post-go-live disruption, and it has surfaced three separate times on one project in three different forms.

The questions to answer, per market

1. Who sends? Which named person or people, in each market.
2. From what address? A personal mailbox, or a shared or generic address such as a market press inbox?
3. If generic, does your identity policy permit it? Many groups require every mailbox to sit behind SSO with an attributable individual identity. An unattributed shared address may not be allowed, regardless of what the client wants.
4. Is everyone on one tenant? Affiliate agencies, partner agencies and recently acquired businesses frequently are not.
5. Is every mailbox migrated to your current mail platform? Not-yet-migrated mailboxes can behave inconsistently, for example sending successfully but not receiving.
6. Who grants admin consent, and how long does it take?

On permissions

We request least privilege by default: the minimum scopes needed to send on a user’s behalf, read replies, and keep the connection alive. Anything beyond that is tied to a specific feature, is optional, and comes with written justification you can hand straight to your IT reviewers.

If your IT function pushes back on a scope, ask us. Every scope we request maps to a named feature, and if you do not need the feature you do not need the scope.

The generic-address problem, stated plainly

Clients often ask for market-branded sending addresses. It reads more professional and it is a reasonable ask. Two things to know before you promise it:

• Your own identity policy may prohibit unattributed mailboxes.
• A no-reply address is not a workaround. It removes the reply, which removes response tracking, which removes most of the value of using the platform at all. If someone proposes it as a compromise, that is the point to push back.

Get this decided, with IT in the room, before you tell the client what is possible.

This is the shortest route from nothing to a sent announcement. It has a specific shape, and the shape matters: it starts with one person, not a list.

The activation path

Steps 1 to 6 are administration. Steps 7 and 8 are where rollouts silently fail. Steps 9 and 10 are where people start getting value, and in practice they are run alongside Session 1 of enablement (Stage 6) rather than as separate training.

Step 1 · The Agency Owner

We create the Agency Account and set up a single Agency Owner. Their job is narrow and front-loaded: they decide who holds admin, and they add the Hub Leads. Nothing else routes through them, which is the point. In a large group this is often the MD or an implementation lead; in a smaller agency it is the actual owner of the agency, and they frequently end up being a Hub Lead as well.

Name the Agency Owner once and then stop thinking about them, except when Hub Leads change.

Step 2 · Hub Leads

The Agency Owner then adds the Hub Leads: two or more named people with identical rights, who between them own the implementation. Everything from here on is theirs. Pick people who will still be doing this in six months, and name at least two, because access admin is exactly the thing that blocks a team the week one of them is on leave.

Hub Leads are the people this guidebook is written for. Their reach is wide, so keeping the role to two or three people avoids a queue forming behind one inbox.

Step 3 · Client accounts

Hub Leads create one account per client.

Step 4 · Nested sub-accounts

Beneath each client account, sub-accounts reflecting how that client is structured. Usually one per market. Where a client has several brands, use one market sub-account covering all brands in that market rather than one sub-account per brand. It keeps access, reporting and commercials simple.

Agree the shape before building it. Restructuring once users, lists and history exist is significantly more work than getting it right at the start.

Step 5 · Users, per account

For each sub-account, Hub Leads add the people who work on it, with the right scope. Account Leads usually need access across several sub-accounts, so identify them up front.

To make this quick, have ready for each person: market, name, role, email address, cluster or hub. Roles matter, because enablement is weighted by role.

Step 6 · Invites, and knowing who has arrived

Hub Leads send invite links and can see who has accepted and who has not. Chase from that view, not from memory or a side spreadsheet.

• Keep the list current. People joining an account team after initial setup are the most common cause of “I never got an invite”.
• If an invite does not arrive, tell us and we will issue a direct signup link rather than leaving anyone blocked. Delivery failures happen, usually down to filtering.
• Seat count is visible to Hub Leads, so you can see what you are using.

Step 7 · Every user needs to manually connect their own inbox after they have logged in

Each user has to connect their own inbox themselves, once, from their own login. Nobody can do it on their behalf: it is their mailbox and their consent.

So:

• Make inbox connection an explicit step in your internal comms, not an assumed part of “getting access”
• Treat “invite accepted” and “inbox connected” as two different completion states, and track both
• Verify it before go-live, not on the morning of the first send

Step 8 · Connecting via the right mail provider

When a user connects their inbox they choose a provider. Choosing the wrong one produces a connection that appears to work and then does not send.

This is not hypothetical. It happens most often where an organisation uses one vendor for documents and calendar and a different one for mail. Someone reasonably presses the button matching the brand they see every day, and it is the wrong one.

Step 9 · Each team adds its own journalist lists

Teams should be able to do this themselves from day one. Do not let list building become something that routes through us, or through one person at the hub.

What to show them:

• The CSV template, and that headers must match it exactly. Header mismatches are the single biggest cause of failed imports.
• Bulk upload, so nobody is adding contacts by hand. If someone is typing journalists in one at a time, something has gone wrong and we want to hear about it.
• Segmenting into focused lists rather than one large one: by language, media type, topic or sector. Smaller, sharper lists consistently outperform.
• Splitting bilingual markets by language before first send. Belgium, Switzerland and Canada all need this. If your lists carry language implicitly, for example in the salutation rather than a language column, flag it and we will help derive it.
• The browser plugin, for adding journalists and publications while reading their work. This is how lists stay alive rather than decaying.
• Keeping bounced and outdated addresses updated. Accuracy decays fast, and it is nobody’s job unless you make it someone’s job.
• Distribution sizing. Cap at roughly 70 journalists per release. Relevance beats reach, and the results are measurably better.

Before moving to announcements, Hub Leads should confirm every market has at least one usable, correctly segmented list. A market with no list cannot do step 10, and that is usually discovered the day it matters.

Where lists are not ready, do not wait. See the start-early path in Stage 4.

Step 10 · Each team creates an announcement

Show the end-to-end workflow, in this order:

1. Intake. Capturing the brief. We have structured question sets designed to pull the right information out of whoever wants the announcement, so teams are not drafting from a vague email.
2. Drafting, against the brand and market rules loaded in Stage 4.
3. Approvals. The steps you documented at Stage 7, built as task templates so a team creates the whole correct sequence in one action instead of assembling it each time.
4. Template selection. Whatever you decided at Stage 0 section E, plain text or branded HTML. Templates must be approved before first send, and this is a common critical-path item.
5. Target list selection, from the lists built in step 9.
6. Personalised pitches per journalist rather than one identical mail to everyone.
7. Scheduling and sending, including how large sends are paced.
8. Coverage tracking switched on, so results accrue automatically rather than being chased manually.

What “done” looks like for this stage

• Agency Owner and Hub Leads set up
• Account hierarchy built and agreed
• All users invited
• All invites accepted, confirmed in the admin view
• Every sending user's inbox connected, via the correct provider
• Every market has at least one usable, correctly segmented journalist list
• At least one template approved and rendering correctly
• A test announcement sent, and a reply received back into the platform, for at least one user per market

That last line is the one that would have prevented the most damage on previous rollouts. A market is not ready because its people have logins. It is ready when someone in it has sent a real email and received a real reply.

What we need

• Existing media lists per market, in our CSV template, with column headers matching exactly
• Topics to monitor, per market
• Competitors to monitor, per market
• Brand and style rules: press release layout and structure, per brand
• Local market syntax rules and language preferences
• Spokesperson biographies per market and brand, with social links
• Core assets for the document library
• Your newsroom URL, if you want journalist visits to it tracked

Expect lists to be uneven

In practice, market media lists arrive in different states of completeness and different formats, and some markets will be at nothing when others are done. This is normal. Do not hold the rollout for it.

Bulk upload and segmentation

Uploading and segmenting lists in bulk is a self-serve function, including splitting by language, market, media type or sector. Your team should not need us for routine list work. If they find themselves needing us, tell us, because that is a product gap and we want to know.

Practical guidance that emerges in every rollout

• Use the CSV template exactly. Header mismatches are the main cause of failed imports.
• Segment into focused lists rather than one large one: by language, media type or topic.
• Cap distribution at roughly 70 journalists per release. Relevance beats reach, and results are better.
• Keep bounced addresses updated. List accuracy decays fast.

Two or three days of work that determines whether you can prove anything at the 90 day review.

Capture, before go-live

• Time spent per announcement, end to end
• Number of announcements per market per month
• Current response rate to pitches, if tracked
• Coverage volume per announcement
• Approval cycle time, from brief to client sign-off

Then agree what the 90 day review will measure, and write it down.

One session per stage, about 30 minutes each, run in order across one to two weeks. All recorded.

Each session maps to one stage and drills the workflow that stage depends on, in the order the stages run. Stage 08 sets out all eight and what each one covers.

Refreshers follow, once people have used the platform on live work rather than in a session. Those are the ones worth booking for a sub-group rather than the whole team.

Sessions build on each other, so run them in order and leave a couple of days between them so people can try things.

Make it role-based

An Account Lead, an automotive media lead and an influencer lead need different things. Tell us your role mix and we will weight the sessions accordingly rather than delivering one generic walkthrough to everyone.

Practical points

• Record everything and index it by topic. With 30-plus people across multiple markets and timezones, some will always miss a session.
• We clip the recordings into short topic-specific segments afterwards.
• The knowledge hub is the persistent self-serve layer. Point people at it constantly.
• Forward the invitations. Whoever built the user list will have missed someone.
• Ask people to submit questions in advance so sessions can be weighted to real concerns.

Baseline first, then sub-groups

The eight sessions exist to get everyone to the baseline: comfortable and confident running an announcement end to end. Treat that as the pass mark for the whole team before anyone specialises.

Then let sub-groups form around real requirements, and tell us what they are. We run smaller sessions with each group in its area, and they become the people their colleagues ask first.

Do this early, not when adoption stalls. Waiting until month four wastes the gap, and it is a much harder sell once people have settled into partial use.

Two things to set up at the same time:

• Pair every enthusiast with a sceptic.
• Name the Decider, so exploration has Account and Hub leads deciding what is worth pursuing.

See the methodology section at the top of this guidebook for why it is structured this way.

Map your approval workflows first

Before go-live, document how announcements move. Typically there are two shapes.

1. Localisation of a global or regional release. Global or regional announcement arrives, gets localised, local market reviews and adjusts, client country lead approves, local market distributes.
2. Local market only. Client country lead briefs the local team, local team drafts, content is reviewed and amended, client country lead approves, local market distributes.

Give us these and we will build task templates that mirror them, so your teams create the right sequence of steps in one action rather than assembling it each time.

Also tell us whether country leads need to approve target lists and supporting collateral, not just the release. That is an approval step people forget to model.

Agency first, client later

Onboard your own teams first. Get them comfortable. Then decide about client access, at Stage 10.

Stage 06 sets the shape of the sessions: one per stage, about 30 minutes each, run in order, recorded and clipped by topic. This is what is inside them.

Each session is built around a workflow rather than a menu of features, and they run in the order below because that is the order in which things have gone wrong on previous rollouts. The first is for Hub Leads only: nothing else can start until the right people have working accounts. The seven that follow are for everyone who sends.

Follow up training tailored to specific needs

These are booked on request rather than scheduled up front, because they land far better with people who already send confidently.

• The AI assistant and the client knowledge base: uploading context, spokesperson biographies, retrieving past correspondence, asking in one language about documents held in another, and editing the custom prompts
• Task templates that mirror the approval workflows you mapped at Stage 07, so the right sequence of steps is created in one action
• The browser plugin for adding journalists and publications while you read, and writing your own pre-send quality checks

Support channel on day one

Set up a shared channel between your team and ours before enablement starts, not after problems appear. Email does not work for this: questions arrive in parallel per-market threads, nobody has visibility, the same issue gets solved three times, and your Hub Leads cannot see what their teams are struggling with.

A shared channel gives market teams direct access to us, gives your Hub Leads sight of the real friction, and gives us the signal we need to fix things. If your organisation needs to add us as external guests, start that request early, because it can take longer than expected.

Fixed review cadence

30, 60 and 90 days, with dates in diaries before go-live. Refreshers promised as “monthly, ad hoc” do not happen. Each review covers usage by market, what is blocked, what people are working around, and progress against the Stage 5 baseline.

Run monthly refresher sessions on top, focused on whatever the reviews surface rather than a repeat of enablement. Later sessions in practice cover the things people only care about once they are using the platform properly: tagging and categorisation, list building at scale, advanced distribution, clearing drafts.

Measurement check-ins

Separate from the adoption reviews above, and arguably more valuable over time. A recurring session, per account, where we bring that account’s own MVPR data and work through it with the team.

It does two jobs at once, and the second matters more as time goes on:

1. Inform what you do next. The next quarter’s plan should be shaped by what your own data says, not by habit.
2. Teach the teams to read it themselves. The goal is that they change what they do because of the numbers, without waiting for us to point at them.

Cadence: monthly per account by default, quarterly as the floor.

What we bring, per account

• Volume: announcements and pitches sent, by market
• Response rate, broken down by market, by list and by journalist
• Coverage volume, and specifically which pitches converted into coverage
• Time from send to first reply
• Which journalists respond consistently, and which never do
• Which topics and angles landed, and which did not
• List health: bounce rate, decay, how much of the target list you reach
• Comparison across markets, so markets can learn from each other

The coaching half

This is the part that changes outcomes, and it is why the session is not a report.

• Turn every number into a decision. A poor response rate on a large list is a targeting problem, not a volume problem, and the fix is a smaller sharper list rather than more sending.
• Name the behaviours that move the metric. Tighter targeting, better personalisation, pitching the journalists who reply, dropping the ones who never do, timing of sends.
• Spread what works across markets. When one market’s response rate is double another’s, find out what they do differently and move it. This is consistently the most underused asset in a multi-market rollout, purely because markets never see each other’s numbers.
• Feed it into the client conversation, so results become a strategy input rather than a monthly report.

Expect and welcome product feedback

When the conversation shifts from “how do I do this” to “this filtering does not work the way we need”, that is the signal adoption is real. Route it to us. Feature requests from live use are the most valuable thing you can send.

Onboarding your client into the platform alongside you is where a lot of the value sits, mainly through faster approvals and shared visibility of results. It is also a bigger step than onboarding your own team, and doing it before your team is confident is a mistake.

So put a dated go/no-go at day 60. Decide yes, no, or not until a specific date, and record which. If your client asks for access before then, that is a strong signal and worth acting on rather than parking.

If you go ahead, your sub-groups can usually run client onboarding themselves by that point, with us supporting.

The maintenance nobody plans for. Most of it is small, and it is what separates an account that is still going strong at month twelve from one that quietly decayed.

Joiners and leavers

Your Hub Leads handle both without coming to us. For every joiner, three things have to happen, not one:

1. Added to the right account and sub-accounts
2. Invite accepted
3. Inbox connected, via the correct provider

Point three is the one that gets forgotten for new joiners exactly as it does at rollout. Keep the one-line connect instruction for your organisation somewhere Hub Leads can paste it into every welcome message.

For leavers: remove access, and remember their connected inbox goes with them. If they were the connected sender for a market, someone else needs to become it, or that market loses its sending route.

Watch for inbox connections dropping

Mail connections can lapse without anyone doing anything wrong: password changes, MFA resets, a policy change on your tenant, or a token expiring. A user whose inbox has silently disconnected looks normal until they try to send.

• Check connection status as a habit, not only when something breaks
• Make it the first diagnostic whenever a send behaves oddly
• After any tenant-wide identity or mail change on your side, assume connections need re-establishing and verify a send per market

List hygiene

Assign an owner per market and a monthly rhythm. Bounces cleared, moves and beat changes updated, dead contacts removed. Lists decay fast, and a stale list quietly reduces response rates long before anyone notices.

Adding new clients and new markets

This is how you scale without waiting on us. Your Hub Leads repeat the Stage 3 activation path for each addition: new sub-account, users, inboxes, lists, test announcement. The same ten steps, and the same rule at the end. It is not ready until someone has sent a real email and had a real reply.

Re-run the Phase 1 questions when your estate changes

A tenant migration, an acquisition, a new affiliate, or a change of mail platform invalidates the answers you gave in Phase 1. When any of those happen, tell us, and revisit question groups A and B. It is a ten-minute conversation that prevents a repeat of the original identity problem.

Seats and commercials

Seat count is visible to your Hub Leads. Review it at your commercial checkpoints so what you are paying for matches what you are using, in both directions.

Feeding product feedback back

Route it through the shared channel. Requests that come from live use are the most valuable input we get, and they are the main reason the platform changes in the direction you need.

Cadence summary

Print this. It gathers the exit criteria from each phase into one place, so it repeats what is already stated in context above.

Phase 1 exit: before enablement is scheduled

• Stage 0 questions worked through, with every unknown assigned an owner and a date
• Sponsoring Account Lead named
• Agency Owner named, and at least two Hub Leads
• Security, AI governance and data protection review paths identified, with owners
• End client’s own tool-approval process checked
• IT Administrator named
• Account Leads identified per cluster or region
• Account hierarchy agreed: client accounts, and sub-accounts per market
• Mail platform confirmed per entity, including affiliates and partner agencies
• Hybrid setups identified (Google Workspace plus Outlook, or the reverse)
• Mid-migration mailboxes identified, with migration dates
• Mailbox access method confirmed per entity, modern API or legacy protocol
• Send rate limits checked on any shared or distribution mailbox
• Sending identity decided per market: who sends, from what address
• Identity policy confirmed to permit that address type
• Tenant audit done: all users, including affiliates and partner agencies
• Admin consent route and turnaround confirmed
• Owner identified for the individual user-consent queue
• Absence cover agreed: who else can send from a named sender’s address
• Media database licence terms checked for export restrictions
• Plain text versus branded HTML decided, per market
• Template variants scoped, and sign-off owner and lead time confirmed
• Salutation and honorific requirements confirmed per market
• Market-specific outbound email compliance requirements confirmed
• External-guest request for the shared channel already submitted
• User list supplied: market, name, role, email, cluster
• Accounts and agency overview live
• Baseline metrics captured
• Shared support channel live
• Sub-groups forming around real requirements, plus a paired sceptic each, plus the Decider
• Role mix supplied so sessions can be weighted
• Internal framing prepared by your champion
• Recording permission confirmed

Phase 2 exit: before go-live

• Target client moment chosen and communicated
• Approval workflows documented and task templates built
• Templates approved by the client where needed
• Contact data loaded, or topic and competitor monitoring running instead
• Topics and competitors supplied per market
• Brand and style rules, market syntax preferences supplied
• Spokesperson biographies loaded
• Invite acceptance and inbox connection both confirmed for every user in the admin view
• Every market has at least one usable, correctly segmented journalist list
• Test announcement sent and a reply received back into the platform, per market
• Mail connections tested end to end, send and receive, per market
• Reviews at 30, 60 and 90 days in diaries
• Day 60 client-onboarding decision in diaries

Phase 3: the ongoing rhythm

• Joiner process includes account, invite and inbox connection
• Leaver process reassigns any market where they were the connected sender
• Connection status checked as a habit, and checked first when a send misbehaves
• List hygiene owner and monthly rhythm per market
• Hub Leads confident running the activation path for new clients and markets
• Agreement to re-run Phase 1 question groups A and B whenever your mail or identity estate changes
• Seat review at commercial checkpoints
• Quarterly review in the diary beyond day 90