Early accessSome features, including AI drafting, are still being switched on.See what works today

Product

How Sendalto works, end to end.

One workspace for lifecycle email and B2B outreach. Plans become campaigns and journeys, every send passes an approval tied to the exact version, and the system enforces your limits at send time.

Example of a campaign waiting for approval: 1,248 trial users, 37 excluded with reasons, three emails, an estimated cost within the monthly cap, and buttons to approve or request changes. Sendalto interface · sample data

1 · Workspace

Start with what is true about your brand.

Each company gets its own workspace: contacts, sending domain, team and brand profile.

A brand profile you confirm

Product, audience, tone, approved facts and the claims you never want made. Every email is written against it, and unknown facts stay unknown.

  • Drafted from one page of your site, then confirmed by you
  • Editable at any time
  • Contacts arrive by CSV import, the Events API or HubSpot

Switching on soon AI drafting of the brand profile is switching on soon. Until then you fill the profile in yourself.

Example brand profile with product, tone, approved facts and a forbidden claim, confirmed by the workspace owner.

2 · Plan

A plan becomes campaigns and journeys.

One-off campaigns for announcements; journeys for anything that should react to what people do in your product.

Readable rules, not opaque scores

A plan lists the audience as a rule you can read, the emails, the waits and the event that ends the journey.

  • AI plan drafts are switching on soon; you can build plans by hand today
  • Every step stays editable
Example AI plan: a segment written as a readable rule, three emails with waits, and an exit when the user starts a subscription.

Journeys that know when to stop

Journeys start from product events and stop when a person reaches the goal or replies. Each run shows why it is waiting, finished or stopped.

  • Triggered by your product events
  • Stop on goal, stop on reply
  • Re-entry rules per journey
Example journey runs: one person completed after reaching the goal, one stopped because they replied and a person now reviews the reply, one is waiting for the next email. Sendalto interface · sample data

3 · Approval

Approval covers this exact version.

Audience, exclusions with reasons, every email and the estimated cost on one screen.

Any edit withdraws it

Change a subject line after approval and the approval is withdrawn; anything already queued from the old version is cancelled.

  • Nothing reaches people unapproved
  • The history shows every version
Example approval history: version 2 approved, then edited into version 3 so the approval was withdrawn, and version 3 now waits for approval.

4 · Sending

Rules checked at send time, not just at planning.

What was true when you approved may not be true an hour later. So each email is checked again right before it goes.

  • Consent and opt-outs

    Consent, unsubscribes, the do-not-contact list and bounces are re-checked for every recipient at send time.

  • Hard monthly spend cap

    Sends reserve budget before they go. When the cap is reached, sending stops; nothing is bought automatically.

  • Two separate sending paths

    Lifecycle email goes through Amazon SES from your verified domain; outreach goes from your own mailbox. Both share one do-not-contact list.

  • One-click unsubscribe

    Marketing email carries one-click unsubscribe headers, and an opt-out is honoured everywhere in the workspace.

5 · Inbox

A reply is a person’s job.

Replies land in a shared inbox, and the sequence pauses for that contact until someone on your team has reviewed it.

Sorted by rules, answered by people

Incoming mail is sorted by rules into opt-outs, out-of-office replies and messages that need a person. Opt-outs are applied automatically.

  • Sequence paused on reply
  • Rule-based sorting, not AI
  • AI reply drafts are switching on soon — and will never be sent by the AI
Example inbox thread: a reply to an outreach email is marked “Needs a person”, the sequence is paused for this contact, and an AI reply draft is marked as switching on soon. Sendalto interface · sample data

6 · Insights

Numbers that don’t flatter.

Insights report what providers confirmed and what people did, and say so when the data cannot answer the question.

Experiments with a control group

Hold back a control group to see whether emails changed behaviour. With too few people, the result says “inconclusive” instead of naming a winner.

  • Control group per experiment
  • Small samples stay inconclusive
Example experiment: 412 people in the treatment group, 398 in the control group, 2,000 required per group, so the result is inconclusive. Sendalto interface · sample data
  • Delivery states as reported

    Accepted, delivered, suppressed or cancelled — each with the reason. “Delivered” never means “in the inbox”.

  • Opens and clicks labelled raw

    Open and click counts are shown as raw, unfiltered signals, because privacy proxies and scanners inflate them.

  • Goals over opens

    Journeys are measured by the product event they aim for, such as a started subscription.

Team

Seven roles, fixed permissions.

Everyone gets the access their job needs, and the server checks every action against the role.

RoleCan
OwnerEverything, including billing, privacy requests, team members, API keys and autonomy settings
AdminEverything except billing and privacy requests
MarketerManage contacts and create campaigns and journeys; cannot approve or send
ApproverReview and approve packages; read content, contacts and reports
SalesWork the inbox for their own accounts and assigned threads
AnalystRead content and reports
TechnicalSet up and verify sending domains

Owners and admins need two-factor authentication and a recent sign-in for sensitive actions such as sending, exports and API keys. Sendalto support sees workspace data only with the owner’s time-limited grant, and every read is logged.

Autonomy

How much the AI may do is the owner’s call.

Three modes. In none of them does anything launch on its own.

  • SupervisedDefault

    The AI only proposes. People create, approve and send everything.

  • Assisted

    AI proposals can be imported as drafts. People still approve and send.

  • Bounded autonomy

    The owner can approve and schedule an AI plan in one step — only within pre-set limits on audience, content, sender, volume and budget.

Set up a workspace and see it for yourself.

Setup and test sends are free. Nothing is billed today, and you see the price before any charge.