Salesforce Nonprofits Landing Page: What Changed and Why

Andreas review, June 2026. Two rounds of changes to the Generic landing page. Round 1 is the launch-prep cleanup (shipped as v1). Round 2 is a proposed restructure (v2) to review before it replaces v1.

Overall: the messaging is great, and the core pitch is strong. Everything below is tightening, not rethinking.

Round 1: launch prep (already applied)

De-identification finish:

The page body was already mostly anonymized
Excluded the assets/clients/ images from the deployed repo (not referenced by the Generic page).
Confirmed the archival diagram PNG contains no client names.

Style cleanup: Removed em dashes per Civis house style. The page had 49 of them; all replaced with commas, colons, or restructured sentences.

Made it standalone + integrated with the main site:

The nav logo pointed at the internal overview page (index.html); it now links to civisanalytics.com, using the same header logo as the main site, with a "Civis for Salesforce Nonprofits" crumb (same pattern as the Audience Builder microsite).
Footer "Company" links were placeholders (href="#"); they now go to real civisanalytics.com pages, plus a legal/social row matching the microsite. "Applied Data Science" was dropped because there's no public URL to point it at; easy to restore if one exists.

Responsive fixes (pre-existing):

The page scrolled horizontally on phones. Two causes: the stack-diagram rows kept a fixed label column at all widths, and the nav buttons plus gutters exceeded narrow viewports. Fixed with media queries; clean now from 360px up.
The hero h1's inline font-size:52px was overriding the 46px mobile rule, so the intended mobile sizing never applied. Fixed.
Added a mobile hamburger menu (the section links previously just disappeared below 1000px). Same pattern as the microsite.
The form's Name/Role fields now stack on phones instead of squeezing side by side.

Round 2: proposed restructure (v2)

A copy review focused on conversion drove three structural changes:

The hero's Reason 02 card had Reason 03's body. Both cards said Claude gives "non-technical marketing and fundraising staff" answers "in plain language," so the three-reason frame collapsed into two. Reason 02 now carries the warehouse/analytics story: "A true data warehouse and analytics workbench behind your CRM: unified data, real reporting, and predictive models your team can act on." Labels sharpened too (Cut storage costs / A real analytics layer / AI for everyone).
The savings story was told three separate times (archival callout, the before/after bars three sections later, the 31% stat ring two sections after that), interleaved with other content. v2 merges all the evidence into one archival section: 50% callout, then before/after bars, then the 100%-accessible and nightly-sync stats, then the flow diagram. The standalone stat-ring section is gone; its 31% number still appears in the before/after bars and case study 1.
Four consecutive Claude sections repeated the same points ("governed" appeared 10 times on the page, "plain language" 5, "MCP" 9). v2 keeps the readiness ladder, the whole-team section, and the stack diagram, and folds the unique content of "Claude in practice" into them (context-freshness became a governance chip; the "one analyst's spreadsheet" line moved into the publishing pillar). The "Why Civis" grid at the bottom of the stack section also repeated earlier material; its two unique points (flat-fee nonprofit connectors, federated chapters/affiliates) replaced the two feature-grid cards that duplicated other sections ("Salesforce archival" and "Claude, built in").

Other v2 changes:

The Salesforce Trends stat band (34/33/31%) moved up to follow "The challenge" section, earlier in the page where it does the motivating; it also no longer sits next to the CTA, where three numbers near 31% could blur with the storage stat.
Section eyebrows now carry "Reason 01 / 02 / 03" so the hero's promise pays off as you scroll.
"Your archival savings alone will nearly pay for the Civis Platform" softened to "For many nonprofits, the storage savings alone offset much of the platform's cost. The rest is upside: a true analytics workbench and AI tools for your entire team." (The original read as a pricing guarantee.)
Case study 1's challenge paragraph repeated "The challenge" section nearly verbatim; rewritten to be case-specific.
"See the Civis Platform running against a data model like yours" appeared three times (CTA, form, modal); now varied.
"We worry about security so you don't have to" was inside the form's consent line; removed (security reassurance already sits in the chips above).
Net effect: three fewer sections, roughly 14% shorter page, no message lost.

Emails: what we changed

All four arcs and their CTAs are kept. The copy was restructured for a cold audience and rebuilt to show how they would look in Pardot, so the renders in this package are exactly what recipients would receive. This is the current template, but we can update the style elements to match closer to the original Claude designs. They exist in Pardot as templates; nothing is scheduled or sent.

Across all four: they now open with a greeting and short, problem-led paragraphs instead of headline blocks, one CTA each, a standardized "The Civis Team" sign-off, the Civis footer, and no em dashes.

Arc 1 (storage): kept nearly as-is; it's the strongest of the set. The closing Claude teaser moved to a P.S. so the email stays one message.
Arc 2 (analytics): rewritten around one concrete outcome: "Your CRM can tell you who gave last year. It can't tell you who's most likely to give again." The tool list was compressed to one reassurance bullet; the full tool wall might read long depending on who receives this one. Can add back in if it's mostly analyst facing.
Arc 3 (Claude): now leads with the lapsed-donor question, the best line in the campaign. The governance paragraph moved after the demo moment, where it answers the objection it raises.
Arc 4 (three reasons): works as the recap; its CTA escalates to Book a demo.

Sequencing (open option): these can run independently as designed, or as a 4-touch sequence (1, then 3, then 2, then 4) where each touch is a new angle. If we sequence, splitting the first touch between Arc 1 and Arc 3 would tell us whether cost savings or AI curiosity converts better with this audience, which then informs everything downstream (including which LinkedIn angle gets budget).

Open items

"Up to 50%" (hero/archival) vs roughly 31% (case evidence): both kept; one clarifying line could reconcile them.
Email sequencing: independent sends vs the 4-touch sequence above.
The form isn't wired up yet; easy to do once we ship it.