How an agency can run client sites from one dashboard
Workflow
September 25, 2026
The short answer
Put each client property you are responsible for on one DashCanopy dashboard, on a plan whose site cap matches the roster. Agency is the tier built for that shape: unlimited sites, health checks every minute, REST API access, and the Pro features under it, including webhooks and SMS through your own Twilio account. Starter and Pro have caps of 5 and 15 sites. If the roster is larger than that, Agency is the plan that matches the pricing page. Every plan is a single login. It is $149 per month, or $1,490 per year.
The dashboard does not become the client’s host, their analytics suite, or their merchant account. It shows whether the sites you added answered, and what stats those sites report.
Decide what “responsible for” means before you add tiles
An agency can have three different relationships with a URL. You host it and you deploy it. You do not host it, but you are on the hook when it is down. You only built it once and you are not on the hook. Monitor the first two. Leave the third off the dashboard unless a contract says otherwise. A red tile for a site you will not repair teaches the team to ignore red tiles, including the ones under contract.
Write the rule in a sentence you can repeat: “If we would hear about it from the client, it is on the dashboard.” Then make the list. A client with a marketing site and a store is two sites. Count sites, not logos. Fifteen logos with two hosts each do not fit on Pro.
Names should include the client. “Homepage” does not work when twelve homepages are in one sidebar. “Northwind — store” does. The description field can hold the internal account name or the person who deploys it. That note is for your team, not for the public internet.
One login today
Every plan is a single login. One person owns the account and the others get the summary that person writes. Do not share the password around the team; you cannot tell who changed a URL, and you cannot remove access without changing it for everyone.
DashCanopy does not give each client a private portal with only their sites. Do not promise a per-client login wall you have not seen in the product.
The REST API, also Agency, is how you pull the dashboard into a report you already generate. It is not required to do the daily read. People can open the tiles. Use the API when a weekly client email should include the same status you see, and you are willing to build that email. Until then, the three lines in the weekly review are enough.
A morning routine a team can share
One person owns the pass on a given day. They read health across every client tile, then stats, then alerts. The order is the same as the morning check. The difference in an agency is what happens after a tile is offline: the owner of the day tells the person who can deploy that client, instead of opening twelve host accounts they do not have.
Keep host logins in the host’s own access controls. DashCanopy’s optional dashboard URL on a site is a bookmark into a console you already can open. It is not a place to store the client’s root password. Do not put client secrets in the site description.
SMS, if you use it, should reach the person on duty, through your Twilio account, on Pro or Agency. A text to a shared phone in a drawer is not on duty. Email should go to an inbox the duty person watches. Rotate the duty, not the monitoring rules.
Checks every minute on Agency mean a recovered site can turn online while you are still writing the client. Look at the tile again before you send “still down.” A Starter interval of 10 minutes is usually the wrong fit for a roster you are paid to watch, both because of the cap and because of the wait.
Client reporting without inventing a second product
A weekly note per active client can be short. Which URLs you monitor. Whether any were offline. Whether reported visitors or revenue, if you connected them, broke their pattern. What you did. That note is yours to write. DashCanopy does not generate a branded PDF from this article’s instructions, and you should not tell a client it does.
Stats only go in the note when the site reports them. A WordPress property can use the plugin. Another platform needs a path that platform allows. Say “not connected” rather than showing a zero that means “we never set up the pipe.” Clients read zero as “we had no visitors,” which may be false. The setup difference is in visitor and revenue numbers on a tile.
When a tile is offline, the sequence in what to do when a tile goes offline still applies. Your extra step is contractual: tell the client what you confirmed, not a guess. “The check failed, we cannot load the URL, we have opened the host” is a factual update. “The server exploded” is not, unless the host said so.
Billing the tool versus billing the client
Your DashCanopy subscription is yours. Manage it from Manage billing on the pricing page or Settings → Payments, which opens Stripe’s customer portal. Canceling or changing the plan is covered in manage or cancel billing. Do not mix that card with a client’s hosting invoice.
How you bill clients for care is your commercial choice. This article is not legal or accounting advice about resale, markups, or agency agreements. Read your own contracts for whether monitoring is included. The product fact is only this: Agency is the published plan whose cap fits a roster bigger than fifteen sites.
Questions agencies ask
Can each client log into only their sites?
Do not promise that from this article. The published Agency features are unlimited sites, one-minute checks, and API access, on top of the Pro feature set. Every plan is one login today, so there is no client-only view.
Is a 20-site roster a Pro plan?
No. Pro stops at 15 sites. Twenty sites is Agency, unless you are willing to leave five unmonitored. Leaving them off should be a choice you can explain, not an accident of the cap.
Should client staging sites be on the dashboard?
Only if you are responsible for them and the team can tell staging from production in the name. A permanent red staging tile, behind a password the checker cannot pass, will drown a real production failure. Prefer production URLs.