Ellis Agency.
Guilford College · Website & Analytics Refresh

GA4 Dashboard & Automated Reports

What this is: a design preview of the reporting you'll receive  ·  Built from: the measurement plan, event taxonomy, GTM implementation plan and the Thriving Together KPI mapping  ·  Prepared by: Ellis Agency  ·  Date: July 2026

How to read this document. These are layout previews — the real thing is built in Looker Studio on live GA4 data. Every panel is labelled so you always know what you're looking at:

Actual  Real figures from Guilford's GA4 property 359308554, pulled 2026-06-22 for the trailing 90 days. These are the pre-launch benchmarks we will measure lift against.

Sample  Illustrative placeholders showing shape and layout only. These panels report on events that do not exist yet — they are created by this project. Giving in particular has zero tracking today, so its first 90 days post-launch become its own baseline. Do not read these as forecasts.

1. How the reporting gets made

One instrumentation layer feeds everything. We tag the shared components, not individual pages, so a CTA emits the same event with the same parameters no matter which template renders it — including the non-Drupal systems, via the portable T6 frame.

1 · Site
Component kit
CTAs, cards, filters, forms push a uniform dataLayer object.
2 · Tagging
GTM container
One container maps dataLayer → GA4. One mapping, all templates.
3 · Data
GA4 359308554
Key events, custom dimensions, audiences, cross-domain.
4 · Reporting
Looker Studio
Dashboard + scheduled email delivery.

2. Screen 1 — the Thriving Together board view

The one page built for trustees. It reports in the Board's language, using the Board's own Measuring Progress targets, and shows the site behaviour that moves each one. Most Board targets are multi-year institutional endpoints — the site reports the leading indicators that ladder toward them, which is why each card names both.

lookerstudio.google.com/…/guilford-thriving-together
Thriving TogetherEnrollmentGiving Audience reachContentWayfindingMobile

Strategic-plan KPIs — pillar 5, “Telling Our Story”

Last 28 days vs prior · refreshed weekly  Sample
Enrollment yield & growth
FY2030 · 1,200 students · $23.5K net/student
applications started
Prospective journey: program pages → Request Info / Visit → Apply start → Apply submit.
apply_start · apply_submit · request_info_submit · visit_scheduled
Geographic diversification
FY2030 · out-of-state >40%
out-of-state share of inquiries
Measured on conversions, not just traffic — region/metro applied to admit-intent actions.
Audience: Out-of-state prospects · geo dimension on key events
Student success & retention
FY2030 · ≥80% fall-to-fall
task-success proxy
The site can't retain students by itself; it removes friction. Leading indicators only — the retention % itself comes from the SIS and is shown for context, not attributed to the site.
Current-student engagement · search success · returning-user engagement
Advancement & campaign readiness
Campaign · ≥$75M by 6/30/2030
gifts started online
The giving funnel and the prominent new Give CTA, with gift value attributed back to its on-site origin.
donation_startdonation_complete {value}

Cross-cut — media visibility & reputation

The plan's eighth outcome

Organic + branded search, direct traffic, and news/referral engagement — the same signals the parallel branding project moves. Sits on this page so trustees see reputation next to enrollment and giving.

Why the KPI numbers are blank above. Three of these four cards report on key events that don't exist yet — they're created by this project. Rather than print invented figures on a board-facing page, the preview shows the structure. The first populated version ships with the post-launch data.

3. Screen 2 — Enrollment

The primary funnel. Today this is not computable: three overlapping vendor events (guilford_application_button, guilford_app_click_to_apply_site, guilford_college_application) measure overlapping things, so a true start→submit rate can't be derived. Standardising the taxonomy is what makes this screen possible.

lookerstudio.google.com/…/enrollment
Thriving TogetherEnrollmentGiving Audience reachContentWayfindingMobile

Admissions funnel

Last 28 days  Sample
Program / hub engagement  program_interest
18,400
↓ 24.5% continue
Admissions hub  page_type:'hub'
4,510
↓ 41.9% continue
Apply CTA clicked  cta_click {apply}
1,890
↓ 66.1% continue
Application started  apply_start
1,249
↓ 47.2% continue  ·  the number nobody can see today
Application submitted  apply_submit
590

Supporting conversions

Sample
Request info
742
▲ 8.4% WoW
true form submits, not clicks
Visits scheduled
168
▲ 12.1% WoW
net-new conversion
Apply start → submit
47.2%
— first measurable
not computable pre-launch
Mobile share of applies
38.6%
▲ 3.2 pts
validates mobile-first build

4. Screen 3 — Giving (net-new)

Donors are the #2 audience and there is no giving tracking at all today — the single biggest measurement gap in the baseline. Giving happens off-domain, so this screen depends on cross-domain measurement and the T6 frame injecting the same tag on the giving platform.

lookerstudio.google.com/…/giving
Thriving TogetherEnrollmentGiving Audience reachContentWayfindingMobile

Giving funnel & revenue

Cross-domain to the giving platform  Sample
Give CTA clicks
2,140
▲ 19.0% WoW
from the prominent new CTA
Gifts started
312
▲ 6.8%
donation_start
Gifts completed
148
▲ 4.2%
donation_complete
Online gift revenue
$41,280
▲ 11.5%
avg gift $279
Give CTA clicked  cta_click {give}
2,140
↓ 38.3% reach the giving platform
Giving landing (cross-domain)
820
↓ 38.0% begin a gift
Gift started  donation_start
312
↓ 47.4% complete
Gift completed  donation_complete {value}
148
Because nothing is tracked today, there is no pre-launch benchmark to beat here. The first 90 days after launch establish the giving baseline, and the Board view reports growth from that point forward.

5. Screen 4 — Audience reach Actual data

This screen is populated with Guilford's real numbers, because it reports on traffic that already exists. It is also where the most important finding in the baseline lives.

lookerstudio.google.com/…/audience-reach
Thriving TogetherEnrollmentGiving Audience reachContentWayfindingMobile

Reach & quality

Trailing 90 days · property 359308554  Actual
Sessions
120,036
90-day baseline
Total users
78,234
72,013 new
Engagement rate
45.0%
avg session 2:34
Key-event conversions
7,044
100% enrollment-flavoured

Channel mix — sessions

Direct highlighted: the “people who already know Guilford” signal  Actual
Organic Search 62,216 · 51.8%
Direct 37,507 · 31.2%
Display (Encoura) 6,113 · 5.1%
Referral 4,856 · 4.0%
Organic Social 3,721 · 3.1%
Paid Other (Encoura) 3,080 · 2.6%
Unassigned 2,977 · 2.5%

Device

Mobile highlighted — the case for the mobile-first rebuild  Actual
Desktop 76,683 · 63.9%
Mobile 42,393 · 35.3%
Tablet 1,351 · 1.1%
The finding this screen exists to protect. Direct is 31.2% and returning sessions are 31.2% — they overlap, so they don't simply add. But the behaviour is unmistakable: between direct, branded organic and returning visitors, roughly half or more of all traffic is people who already know Guilford — current students, faculty/staff and applicants checking status. Top pages bear it out: /current-students, /faculty-and-staff, /account/login, /admitted-students, /apply/status. Do not narrow the site. Those audiences stay first-class in both the IA and the reporting.

6. Screen 5 — Content & wayfinding

lookerstudio.google.com/…/content-wayfinding
Thriving TogetherEnrollmentGiving Audience reachContentWayfindingMobile

Top pages

Views / sessions · trailing 90 days  Actual
PageViewsSessionsPrimary audience
/44,53737,370All
/current-students15,94612,506Current students
/faculty-and-staff8,1936,542Faculty & staff
/admissions/apply7,8407,378Prospective
/directory7,6272,184All
/program-finder6,1794,854Prospective
/admitted-students4,4683,939Admitted
/account/login4,0272,700Current students

Search & task success

Sample
Site searches
1,349
existing view_search_results
Zero-result rate
8.2%
▼ 1.4 pts
new signal
Quick-links CTR
34.6%
▲ 2.0 pts
wayfinding proxy
Returning engagement
52.1%
— guardrail
must not regress

7. The automated weekly report

Delivered by Looker Studio scheduled delivery — no one has to log in or pull anything. This is the email itself, not a link to a dashboard: the numbers that matter are in the message, with the dashboard one click away for anyone who wants to dig.

Guilford weekly — enrollment, giving & reach · week of 20 Jul
From: Guilford Analytics (Looker Studio)  ·  To: Guilford marketing team; Ellis Agency account lead  ·  Mondays, 8:00 ET
Week of 20 July 2026
vs previous week · guilford.edu · GA4 359308554
The week in one line

Applications started rose 8.4% on the back of program-page traffic; giving held flat; mobile share of applies crossed 38% for the first time. Sample

Headline
Sessions
9,842
▲ 5.1%
Apply starts
1,249
▲ 8.4%
Request info
742
▲ 3.2%
Gifts completed
148
— 0.4%
Strategic-plan KPIs
Board KPIThis weekTrend
Enrollment yield & growth1,249 starts▲ 8.4%
Out-of-state share of inquiries36.2%▲ 1.1 pts
Student task success (proxy)91.8%— flat
Advancement — gifts online$41,280▲ 11.5%
Worth a look
Zero-result searches fell to 8.2% after the navigation change — the consolidated menus appear to be doing their job. Worth confirming next week before we call it.
Giving flat for a third week. Give CTA clicks are up 19% but completions aren't following — the drop-off is on the giving platform, not the site. Suggest Advancement reviews the gift form.
Open the full dashboard →
Automated weekly delivery · Looker Studio · Ellis Agency for Guilford College. A daily operational version goes to the Ellis team only. Sample content shown for layout.
Why the email carries commentary. An automated dashboard link gets opened once. The weekly report leads with what changed and what it probably means, so it stays useful to people who don't live in analytics — which is the whole point of "automated reports" in the kickoff brief.

8. What powers it — tagging & GTM

The dashboard is only as good as the taxonomy underneath it. These are the key events the build creates, and what each one replaces.

Key eventFires whenReplaces / status
apply_startAny Apply CTA / handoff to the application portal Replaces guilford_application_button, guilford_app_click_to_apply_site
apply_submitApplication submitted (cross-domain confirmation) Replaces guilford_college_application
request_info_submitRFI form submit Replaces click-proxies — moves from click to true submit
visit_scheduledVisit/tour bookedNew — no equivalent today
donation_startGive flow begins (off-domain)New — biggest gap
donation_completeGift confirmed, with valueNew — biggest gap

Every event carries a shared parameter set, so any number on any screen can be sliced the same way: template, page_type, audience_priority, cta_type, position, device, form_type, program, and value/gift_designation for giving. position is how we will answer “did making Give prominent actually work?”

Continuity matters more than cleanliness. For ~30 days at cutover we fire the legacy vendor events and the new ones in parallel, and keep a mapping table, so the pre-launch baseline stitches cleanly to post-launch. Legacy events are never deleted from history — the cutover date is annotated instead. Without this, the refresh would look like a traffic cliff.

9. Cadence & distribution

ReportCadenceAudienceShape
Weekly reportMondays, 8:00 ET Guilford marketing + Ellis account leadThe email above
Daily operationalEvery morningEllis ops (internal)Anomaly watch — catches tagging breaks fast
Thriving Together board viewMonthly / quarterly roll-up Senior leaders & trusteesScreen 1, in the Board's “Measuring Progress” shape
Live dashboardAlways onAnyone with the link Embedded in the client portal deliverable

10. What we still need to build it

Honest dependencies — these gate the screens above, and most need someone at Guilford: