Analytics for Multiple WordPress Sites: One Dashboard, Honest Numbers
Twenty sites, twenty logins, twenty consent banners and twenty different definitions of a visitor. There is a simpler way to run this.
On this page
Analytics for multiple websites is an overhead problem, not a data problem. Any tool can count visits on twenty sites. The work is in twenty setups, twenty logins, twenty consent banners, and twenty slightly different definitions of what a visitor is. The fix is to run one service that treats sites as rows under one account, applies the same bot rules everywhere, needs no consent banner for analytics, and exposes an API so you can stop opening dashboards one at a time.
What goes wrong at scale
With a single site, you set up analytics once and forget it. With a portfolio, three things compound:
Setup per site. A tag, a property, a consent integration, a set of events. Multiply by twenty and it is a day of work, repeated whenever a site is added or a plugin changes.
Inconsistent numbers. Site A has bot filtering on, site B does not. Site C’s consent banner gets 40 percent acceptance, site D’s gets 70. Comparing their traffic tells you about their configurations, not their audiences.
Reporting by hand. Monthly client reports mean opening each property, screenshotting, pasting. Or building a Looker Studio board that breaks when a property changes.
What a multi-site setup needs
One account, every site inside it. Add a site, get its tracking in place, see it in the same list as the others. DevDome Analytics works this way: one hosted dashboard for unlimited sites under a free DevDome account, with the WordPress plugin registering each site. We make it, so weigh that; the structure is what to look for in any tool.
The same rules on every site. Bot classification, session definitions and click capture should be identical across the portfolio so that a visitor means the same thing everywhere. DevDome scores every hit on six signals and shows the human versus bot split per site and in total, with plans metered on human pageviews only.
Cookieless by default. No analytics cookies and no profiles means no per-site consent banner work for analytics alone, which is the largest single saving across many sites. Not legal advice; anything else that needs consent still needs it.
Server-side click capture. On affiliate and lead-gen portfolios, outbound clicks are the number that matters, and browser blockers remove the script on a share of visits. A server-side hop records the click regardless.
An API. With more than a handful of sites, you want every site’s humans, bots, top pages and clicks in one sheet. A public Stats API, and an MCP server for AI tools, turns the monthly report into a script.
Setting it up on a portfolio
- Create the account once and add sites from it. On WordPress multisite, network-activate the plugin and connect each sub-site, or connect only the ones you report on.
- Standardise the plugin settings. Same tracking options on every site: which clicks count, whether logged-in users are excluded, site search tracking. Write them down.
- Decide reporting periods. The DevDome dashboard offers 24 hours to 12 months. Pick the period for client reports and use the same for everyone.
- Give clients their own view. Read-only, per site, revocable. Your login stays yours.
- Automate the monthly numbers. Pull human pageviews, visitors, top pages and clicks per site through the API into a sheet. Review the sheet, not twenty dashboards.
The consent problem, multiplied
On one site, a consent banner is an afternoon: install the plugin, wire the analytics tag to the consent state, test that nothing fires before acceptance. On twenty sites it is a week, repeated whenever a site adds a plugin that sets its own cookies or a regulator changes its guidance. And the banner costs traffic on every site: a share of visitors decline, and GA4 then measures a fraction of them. Cookieless analytics removes that work for the analytics part. No analytics cookies and no profiles means the analytics line comes off the banner, and on sites that have nothing else needing consent, the banner itself can go. That is not legal advice; anything else on the site that needs consent still needs it. But across a portfolio the saving is the single largest one, and it compounds with every site you add.
Client access without sharing your login
The agency habit of sharing one dashboard login with clients is how a client ends up seeing another client’s numbers, or how a departed client keeps reading your data. The right shape is per-site, read-only access you can revoke: an invite or a share link scoped to one site, with your own account untouched. When the engagement ends, revoke that one share. If a client wants their numbers in their own tools, an API key scoped to reading is the alternative; it lets them pull their site’s reports into a sheet without any access to the dashboard. Either way, the rule is that nobody but you holds the account credentials.
Comparing sites honestly
Once every site runs the same rules, comparisons mean something. Three that are worth doing monthly:
- Bot share per site. A site with a far higher bot share than its siblings is being scraped or probed. Worth a look at its 404 log and login attempts.
- Human pageviews per post published. Sites that publish more do not always earn more; this ratio shows which content strategy is working.
- Click-through on tracked links. For affiliate portfolios, the sites where humans actually click are the ones to invest in.
None of these are possible when each site measures differently.
When to stay with Google Analytics
If your clients run Google Ads and need audiences and attribution, keep GA4 on those sites. You can run both: GA4 for the ads team, a cookieless multi-site dashboard for honest day-to-day numbers and client reporting. The two will disagree, and the gap is bots and blockers, which is itself a useful number to show.
For everyone else, one account, one set of rules and one API is the difference between analytics as a monthly chore and analytics as a glance.
Related reading: ChatGPT Referral Traffic: How to See It in Your Analytics and What It Is Worth.
Disclosure: DevDome publishes this blog and makes the products it mentions. We describe what they do rather than promise results.
Key takeaways
- The cost of multi-site analytics is not the data, it is the overhead: separate properties, logins, consent setups and inconsistent bot filtering.
- Pick a tool that treats sites as rows under one account, with the same classification rules on every site, so numbers are comparable.
- Cookieless tracking removes the per-site consent banner work, which is the biggest time sink across a portfolio.
- An API matters at scale: pull every site's human pageviews into one sheet instead of opening twenty dashboards.
- Client access should be per site, read-only, and revocable without touching your own login.
Sources
- Google Analytics Help: account and property limits : how many properties one account can hold
- WordPress.org: Multisite network administration : how network activation and per-site settings work
- Google Analytics Help: Looker Studio integration : the usual route to a cross-property GA4 view
Links last checked October 17, 2026.
Frequently asked questions
Can I use one Google Analytics account for many sites?
Yes, one account with a property per site, up to Google's limits. The overhead is in the setup: each property needs its own tag, consent configuration and report views, and GA4's interface does not compare properties side by side without Looker Studio or exports.
What about WordPress multisite?
A network is many sites sharing one install. A plugin that registers each sub-site separately under one account gives you per-site numbers and a combined view. Make sure the plugin supports network activation if you want it on every sub-site at once.
How do I give a client access to only their site?
Use a tool with per-site sharing or viewer accounts. Never share your own login. Read-only access per site, with a link or an invite you can revoke, keeps the client informed and your other clients private.
Will bots be counted differently on different sites?
With separate tools or separate configurations, often yes, which makes comparisons meaningless. A single service applying the same classification to every site gives you one definition of a human visitor across the portfolio.