Overview, digest and uptime

The Overview shows every site in your account on one page. Sites that need you come first, each with the one thing to do next.

Each site is a card. Its numbers sit in a row of tiles: Visitors, 7 days, with the last 7 days drawn and the change against a usual week; the goal's conversions, Ad spend, Cost per result and Tracking health when the site has them; and Uptime. Open dashboard opens the site.

Cost per result is ad spend divided by results from your ads: the goal's conversions Datacenter credits to an ad platform that spent, as the Attribution report credits them, over the last 7 complete days. Conversions from search, email or word of mouth cost nothing, so they are left out. The line under it says how many came from your ads. When none are tied to the ads, it divides by the results the ad platforms report, and says so. With neither, the tile is left out.

What counts as needing attention#

Worst first:

• Site down. Two uptime checks in a row failed.
• Tag stopped. No visits in 24 hours, on a site where a usual day brings 10 or more.
• Conversions or visitors fell. The last 7 days came in under 70% of a usual week, and well outside how much the site usually swings. A usual week is the middle of the 4 weeks before. Small sites are left out: under 70 visitors or 5 conversions a week, a drop says little.
• Ads spent with no results. Ads spent in the last 7 days, and the site's goal got nothing. With no goal yet, it asks you to add one, unless the ad platform reports results, like signups from Meta. Those show on the row instead.
• SSL or domain ending soon. The certificate ends within 14 days, or the domain within 30.
• An unusual change. In the last 2 days a number moved far from a usual day for that weekday, the way that hurts: a top source or sales fell, or ad spend or cost per result rose. Muted ones are left out. See Unusual changes.
• A sync stopped. A connected tool, such as Shopify, stopped syncing.
• Tracking health is low. The site's tracking health score is under 60 of 100.
• A report waits for your OK. A report set to Check first is ready for you.

Visits refresh every hour and goals once a day, so the page loads fast however many sites you have. Pick a client name above the list to see only that client's sites.

What's happening on each site#

Under What's happening, each site has a short briefing for the last 7 days: a few plain sentences from its numbers, then its weekly plan and any suggested goals. Open, Not now and Dismiss work right there. A site's dashboard opens straight on its charts.

The digest#

In Alerts, then Settings, under Digest and Teams, choose Every day or Every Monday, the hour and your time zone. The digest lists each site that needs you with its action, then how many are all good. It goes to your email, and to Slack or Teams when you add them. Studio and Agency include it.

Slack

Slack gets the digest, the alerts you pick and, if you like, each site's Monday plan. You make an incoming webhook in your own Slack workspace, so there is no app to install or approve (Slack's guide, checked October 2026).

1. Open api.slack.com/apps, press Create New App, then From scratch. Name it Datacenter, pick your workspace and press Create App.
2. In the menu on the left, open Incoming Webhooks and turn on Activate Incoming Webhooks.
3. Press Add New Webhook to Workspace, pick the channel and press Authorize. For a private channel, join it first.
4. Under Webhook URLs for Your Workspace, press Copy. The address starts with https://hooks.slack.com/services/.
5. In Datacenter, open Alerts, then Settings. In the Slack card, paste it, tick what to post and press Save, then Send a test.

What you can post: Traffic drop, Tracking broken (the tag stopped, or a tracking check failed), Site down, SSL and domain, Big day, Target behind, New sale over an amount you set (in the sale's own currency, from a connected store or Stripe), and Setup request done. Under Also post, tick The digest and Monday plan.

Each post has a title, a line or two, the site's name and buttons into the app. Big day and Target behind also get a Client view button that opens the client's dashboard on your client portal, once it is live.

One site can differ: pick it above the cards, then Choose for this site. Tick its own alerts, or open Post to a different channel to give it its own webhook, for a channel per client.

Quiet hours hold Slack posts as they hold email alerts, set in each site's alert settings. What waited goes as one message when they end. Site down never waits.

If Slack says the webhook no longer exists (it was removed, or the channel was archived), Datacenter stops posting to it and emails you once. Make a new webhook and paste it to turn posts on again.

Microsoft Teams

1. In Teams, go to the channel, select More options (...), then Workflows.
2. Pick the template Send webhook alerts to a channel, set it up and select Save.
3. Copy the webhook link it gives you.
4. Paste it in Alerts, then Settings, under Digest and Teams, and press Save, then Send a test.

Teams gets the digest and, if you tick it, site down, SSL and domain alerts. Microsoft retired Office 365 connector webhooks in May 2026, so only Workflows addresses work (Microsoft Learn, checked October 2026).

The address lets anyone who has it post to that channel, so Datacenter stores it encrypted and shows only its last four characters.

Uptime, SSL and domain checks#

On paid plans, Datacenter loads each site's home page every 15 minutes.

• Each check tries three times before it counts as failed, and a site is called down only after two failed checks in a row. One slow minute never wakes anyone up.
• A site is up when it answers. Server errors (500 and up), "page not found" on the home page and no answer in 15 seconds count as down.
• A site behind a Cloudflare bot challenge answers for Cloudflare, not its server. The Uptime card says so, and how to let the check through.
• The SSL date comes from public certificate logs, looked up once a day. Certificates usually renew a month early, so one that ends within 14 days usually means renewal is failing.
• The domain date comes from the registry, looked up once a day. Some registries do not publish it; then the page says so.

You get an email when a site goes down and when it is back, once each per outage, and when a certificate or domain is about to end: 14 and 3 days before for a certificate, 30 and 7 for a domain. Tick Site down, SSL and domain in the Slack card, or Also post site down, SSL and domain alerts to Teams, to send them there too. In Alerts, tap a site's name to see its uptime for the last 30 days. To stop these emails, turn off Site down, SSL and domain under Emails. Add the Uptime block to a report to show it to your client.

The report summary#

Every client report opens with a short summary: What happened, Why, What we did and What's next, written from that period's numbers.

• What we did lists changes made in Datacenter during the period, such as goals added and email flows set up.
• What's next comes from the site's Autopilot plan.
• With the Assistant on, AI writes it and uses one AI action per report period. Datacenter then checks every number in it against your data. If one does not match, the plain version written from the numbers goes instead.

In the report's Summary card, pick Automatic to send it with each report, Check first to get an email when it is ready and send the report when you approve, or Off. You can change the words in either mode.