Skip to main content

Settings

In plain English

Everything you can configure is here, in fourteen tabs.

You don't need to visit most of them. Four matter on day one:

TabWhy it matters immediately
BudgetCaps your monthly AI spend. Set this first
BrandMakes everything the AI writes sound like you
IntegrationsConnects your other tools. Check for expired connections
TeamWho else can get in, and what they can do

The rest you'll visit when you need them — changing a default, checking your subscription, adjusting notifications.

One tab deserves a warning. The Danger Zone does exactly what it says: permanent, irreversible things. Nothing in there can be undone with a support ticket.

How to use it

Your first fifteen minutes

  1. Click "Settings" in the left-hand menu.
  2. Go to Budget. Set a monthly limit you'd be comfortable seeing on a bill. You can raise it later.
  3. Go to Brand. Fill in your voice, tone, and any words you never want used. This makes every writing tool better.
  4. Go to Integrations. Connect the tools you use, and note which ones show as needing attention.
  5. Go to Team if anyone else works on the shop. See Team & permissions.
  6. Leave everything else until you have a reason to change it.

⏱ ~15 min to set up properly · 💳 Starter+ · 🎯 Configure once, then forget about it

Why this matters for your business

Two settings here account for most of the problems merchants actually hit.

The first is the AI budget. This app uses AI in a lot of places — writing descriptions, generating campaigns, analysing reviews. Each use costs something. Without a cap, a bulk operation across a large catalogue can run up a bill that surprises you at the end of the month. Setting a limit takes thirty seconds and means the worst case is "something stopped" rather than "something cost more than expected."

The second is integrations. Connections to other services expire. Access gets revoked, passwords change, permissions get withdrawn during a tidy-up. When that happens, the thing that connection powered stops working — and usually stops silently. Your campaign doesn't send, your feed doesn't update, and nothing obviously errors.

Checking the integrations tab when something unexpectedly isn't working is one of the highest-return two-minute habits available. It's the answer far more often than people expect.

Beyond those two, this screen is mostly set-and-forget. The value is in getting the first setup right rather than in visiting often.

What this typically unlocks

What you getTypical result
Unexpected AI costsCapped by your budget
Silent failures from expired connectionsDiagnosable in ~2 min
Quality of AI-written copyNoticeably better with Brand filled in
Configuration huntingEliminated — one screen
Accidental destructive actionsGuarded behind a clearly-marked zone

What you actually get

The fourteen tabs

TabWhat it's forVisit it
OverviewA summary of your current configurationOccasionally
AI ProviderWhich AI service powers the writing toolsRarely
DefaultsDefault settings the tools start fromWhen something keeps needing the same change
BudgetYour monthly AI spending capDay one
SubscriptionYour plan, usage, and billingWhen changing plan
BrandVoice, tone, banned phrasesDay one
TeamPeople and their rolesDay one, if you have a team
IntegrationsConnections to other servicesDay one, then when something breaks
BarcodeBarcode and product-identifier settingsIf you use barcodes
NotificationsWhat you're told about, and howEarly
PresetsSaved configurations you reuseWhen you find yourself repeating setup
JobsScheduled background workRarely
SupportHelp and contactWhen you need it
Danger ZonePermanent, irreversible actionsAlmost never

Budget — set this first

Caps what the app can spend on AI per month.

SettingWhat to consider
Monthly limitSomething you'd be comfortable seeing on a bill
Alert thresholdGet warned before you hit the cap

What happens at the limit. AI features stop until the next period or you raise the cap. That's inconvenient — and much better than an unexpected bill.

A sensible starting approach: set it lower than you think you need. Watch actual usage for a month. Raise it if it's genuinely constraining. Merchants who set a high cap "to be safe" are the ones who get surprised.

Brand — the highest-value tab

SettingWhat it does
Voice and toneHow the AI writes — formal, warm, direct
Banned phrasesWords and phrases never to use
Brand assetsLogo, colours, fonts, imagery style

Everything the AI writes uses these. Fill it in once and every tool improves.

Banned phrases are worth thinking about. Competitor names, claims you can't support, words that don't fit your brand. Violations are caught and reported in Brand Audit.

Pair this with Positioning — brand voice controls how things are said, positioning controls what is said. Both together is what makes AI copy sound like you.

Integrations — check here when things break silently

Every connection to an outside service: ad platforms, messaging providers, marketplaces, analytics.

StatusWhat it meansWhat to do
ConnectedWorkingNothing
Needs attentionExpiring or degradedReconnect soon
ExpiredStopped workingReconnect now

Expired connections fail quietly. Your messages don't send, your feeds don't update, and nothing shouts about it. If something unexpectedly isn't happening, check here before anything else — then check Job Platform for the failed work, which can usually be retried once the connection is fixed.

Reconnect promptly when a connection shows "needs attention." That's a warning before the failure, and it's much cheaper to act on than the failure itself.

Notifications

What the app tells you about and where it sends it.

Worth turning onWhy
Critical alertsThings actively going wrong
Budget warningsBefore you hit your AI cap
Integration expiryBefore a connection dies
Weekly summaryThe digest — see Goals & alerts

Send them somewhere you'll see them. A shared team channel beats an inbox that fills up.

Danger Zone

Permanent, irreversible actions — bulk deletions, resets, disconnections that lose history.

Nothing here can be undone. Not by you, not by support.

Before you touch it
Have you exported anything you might need?
Do you understand exactly what it affects?
Is there a less destructive option?
Is this the right shop?

If you're not certain, contact support first. The Danger Zone being clearly marked is the last safety net; use it as one.

How it works (without the technical bits)

Real merchant scenarios

Scenario A — The bill nobody capped

Setup. Merchant with 6,000 products ran a bulk description generation across the whole catalogue, then ran it again with a different tone, then a third time on a subset.

No budget cap was set.

Their AI spend that month was several times any previous month, and it wasn't visible until the bill.

Nothing was broken. The work was legitimate and the output was used. It just cost more than anyone had thought about, because nothing stopped to ask.

Afterwards. Set a monthly cap with an alert at 70%. The next time a large run approached the limit, they got a warning, checked whether it was worth continuing, and decided deliberately.

The cap didn't reduce what they spent overall. It made the spending a decision rather than a discovery.

Scenario B — Two weeks of silent failure

Setup. Apparel brand's WhatsApp campaigns stopped producing results. Sends appeared to be scheduled correctly.

They spent two weeks reviewing content, audiences, and send times.

The actual cause was on the Integrations tab:

WhatsApp Business Expired Reconnect required

The connection had expired 15 days earlier. Every send since had failed.

Fixed in three minutes once found. The failed sends were retried from Job Platform and delivered.

Two weeks of investigation that a two-minute check would have ended on day one.

Their new rule: when something stops working, check Integrations before anything else.

Scenario C — Brand settings changing everything

Setup. Merchant used the AI writing tools for three months with Brand left empty. Output was generic and needed heavy editing every time.

Filled in the Brand tab — tone, style, and a list of banned words including three competitor names and four phrases that didn't fit their positioning.

BeforeAfter
Editing time per piece~15 min~4 min
Pieces needing a full rewrite~30%~5%
Brand-rule violations caughtn/a2 in first month

Twenty minutes of setup, and roughly eleven minutes saved per piece across everything they generated afterwards.

Scenario D — The Danger Zone action that went ahead anyway

Setup. Merchant wanted to clear test products from a development phase. Used a bulk delete in the Danger Zone.

They selected a filter that matched more than intended — it included 40 live products that shared a naming convention with the test items.

Nothing could be recovered. Not by them, not by support.

Rebuilding took three days, and some product review history was permanently lost.

What would have prevented it. Exporting first, and testing the filter on a small selection before running it across everything.

Why this is in the docs. The Danger Zone is clearly labelled, and people use it anyway when they're in a hurry. Treat the label as the warning it is.

Scenario E — Notifications to the right place

Setup. Critical alerts went to the owner's email, which received several hundred messages a day.

A payment-integration failure alert sat unread for six days. Subscription renewals failed throughout.

Action. Routed critical alerts to a shared team channel instead of an individual inbox.

Email onlyShared channel
Average time to notice a critical alert~3 days~20 minutes
Alerts missed entirelySeveral0

Same alerts, different destination. Where a notification lands matters as much as whether it's sent.

Best practices

Set the AI budget on day one, lower than you think you need.

Fill in Brand before generating anything at scale. Scenario C.

Check Integrations first when something breaks silently. Scenario B.

Reconnect on "needs attention," not on "expired."

Send critical alerts to a shared channel. Scenario E.

Export before anything in the Danger Zone.

Don't set a high budget cap "to be safe." That's the opposite of safe.

Don't skip Brand because it's a form. It affects every writing tool.

Don't use the Danger Zone in a hurry. Nothing there is reversible.

Don't route alerts to a busy personal inbox.

Plan tiers

CapabilityFreeStarterProAgencyEnterprise
Core settings and defaults
AI budget controls
Brand voice and banned phrases
Integrations
Team management
Presets
Scheduled jobs configuration
Cross-store settings templates

Frequently asked

What happens when I hit my AI budget? AI features stop until the next period or you raise the limit. Set an alert threshold so you're warned first.

Where do I change my plan? The Subscription tab.

How do I know if an integration has expired? The Integrations tab shows status per connection. Turn on integration-expiry notifications to be warned before it happens.

Does changing Brand settings update existing copy? No. It affects what's generated from that point on. Regenerate anything important.

Can I undo something from the Danger Zone? No. Not by you, not by support. Export first.

Where did the old Billing and Support pages go? Both were consolidated into Settings — Subscription and Support tabs respectively. The old links redirect here.

Who can change settings? Depends on their role. See Team & permissions.

What are Presets for? Saved configurations you reuse. Useful when you repeatedly set up the same options.

See also