Research · The Raw Report

The research underneath the STRATEGY.

The full report behind the positioning, UX and conversion best practices, AI-search strategy, the competitor white space, and the citable market facts that back the USP. Every claim links to its primary source.

← Back to the walkthrough

Website Copy & Information Architecture Playbook for The Sax Agency: Best Practices + Citable Facts (2025–2026)

TL;DR

  • Structure the site around the buyer, not the agency. Use 5–7 short (1–2 word) primary navigation labels with strong “information scent,” a plain-language descriptive H1 (“we do X for Y”), client-focused (“you”) copy, and a hero → value-prop → proof → services → CTA flow. These are documented UX/conversion best practices from Nielsen Norman Group (nngroup.com) and Orbit Media’s review of 500 sites (orbitmedia.com).
  • Win both Google and AI search with research-first depth. Publish comprehensive, people-first content that demonstrates E-E-A-T, use clean JSON-LD schema (Organization, ProfessionalService/LocalBusiness, Service, BreadcrumbList), keep title links under ~600 px and meta descriptions ~150–160 characters, and format content “answer-first”, AI Overviews already appear on roughly one-in-five Google searches and most of their citations come from top-10 organic results, so depth + structure is now a dual SEO/AEO bet.
  • Back the copy with primary-source facts and own the white space competitors ignore. Combined Black/Asian/Native American buying power reached $3.2 trillion in 2021 (Selig Center); Black-owned firms are only ~2.5–3.3% of US businesses (Census); 331 jurisdictions plus 3 states, including all 58 California counties, require language assistance under VRA Section 203 (Census Bureau); and a third of consumers cut spending from brands retreating on DEI (Collage Group). Sax’s public-sector/civic prime-AOR record + measurable impact + language-access expertise is a position no leading multicultural agency currently claims.

Key Findings

The single most important strategic conclusion: Sax should build a “research-first, proof-led, buyer-centric” website that leads with civic/public-sector credibility and measurable impact, because (a) that is exactly what UX, conversion, and AI-search best practices reward, and (b) it is the one positioning territory that Burrell, Walton Isaacson, Egami, and Carol H. Williams have left open. Everything below supports executing that thesis.

Details

1. Website Navigation & Information Architecture (UX)

Keep primary nav labels to 1–2 words and front-load meaning. Nielsen Norman Group’s menu-design guidance is explicit: “Use succinct labels of 1–2 words in length for each navigation link… enhance scannability by starting with the most information-carrying word and avoid made-up terms.” It also warns against vague or branded labels that lack “information scent.”

Why short labels beat long descriptive ones: users scan menus rather than read them; NN/g recommends left-justifying vertical menus and front-loading key terms so users “spend less time reading menus.” Long, conversational, or parallel CTA-style labels create confusion and weaken information scent (NN/g, “vague call-to-action verbs… are confusing. Use labels with strong information scent instead”).

When to use dropdowns / mega-menus: NN/g advises mega-menus when a section contains many pages/subsections (e.g., grouping multiple sector pages and multiple service pages); Orbit Media’s recommendation is to “use dropdowns only if the section has a lot of pages and subsections… avoid little dropdowns with few options because they don’t test well in UX studies.”

Conventions to honor (Orbit Media’s 500-site dataset, mostly B2B): logo top-left on 93% of all websites; horizontal menus that collapse into a hamburger on mobile are “standard and used by 90% of marketing websites.” Two-thirds of homepages lack a real (verb-driven, specific) CTA, a fixable gap for Sax.

Recommended Sax primary nav (5–7 one-word labels): Work · Sectors (mega-menu: Public Sector & Civic / Nonprofit & Foundations / Sports, Entertainment & Culture) · Services (mega-menu) · Impact · About · Insights · Contact. This keeps labels short, frontloads meaning, and uses a mega-menu only where there are genuinely multiple sector/service pages to group.

2. Agency Website Copy & Brand Voice

Client-focused (“you”) copy outperforms self-focused (“we”) copy. The consistent guidance across copywriting/CRO authorities is to make the reader the hero: replace “we are…/our firm…” with the client’s problem and outcome. As one widely cited framing puts it, the difference between “We offer comprehensive solutions” and “You’ll finally solve this frustrating problem” is the difference between informing and converting; client-focused copy “speaks directly to your client’s needs, their problems and how you solve them.” The recommended balance is not zero “we”, it is to open with the client’s world, prove with your record, and reserve “we/our” for the proof and About sections.

Homepage / service-page conversion structure (Orbit Media, 1,000+ sites built):

  1. Logo top-left, restrained.
  2. Descriptive H1, 6–10 words, stating what the company does and including the target keyphrase (“clear beats clever”; pass the “backyard BBQ test”). Note: 54% of homepages put the value proposition in the H1.
  3. Brief intro / value proposition that positions and differentiates in a few lines.
  4. Prominent CTA (specific verb, contrasting color, high on page).
  5. Proof / social proof (clients, testimonials with names + photos, results).
  6. Services, grouped, each with a benefit-led short description.
  7. Supportive data (statistics/case studies) and content (Insights) near the bottom with an email CTA.

There is a measurable gap between what agencies build and what B2B buyers want, Orbit Media + Ascend2’s original research found contributors over-index on brand story/social proof while visitors prioritize clarity, the people behind the company, and reviews/testimonials. Implication for Sax: lead with clarity and named proof, not mission language.

3. Long-Form Content & SEO

Does long-form rank better? Correlation, not causation. Multiple analyses (Backlinko, Ahrefs, Semrush) find top-ranking pages tend to be longer (commonly cited ~1,447–1,890 words on page one), but the cause is comprehensiveness, not word count. Semrush states it plainly: “Google doesn’t rank content based on word count. Instead, it evaluates whether content satisfies user intent. A 1,000-word page can outrank a 3,000-word article if it better aligns with searchers’ needs.” “Depth that rewards the visitor” means covering every sub-question a buyer has, adding original data/case studies, and exceeding competitors with value, not filler.

E-E-A-T and helpful content (primary Google sources): Google’s automated systems “prioritize helpful, reliable information that’s created to benefit people.” Self-assessment questions include whether content provides “original information, reporting, research, or analysis” and “a substantial, complete, or comprehensive description of the topic” with “clear sourcing, evidence of the expertise involved.” “Experience” was added to E-A-T (→ E-E-A-T) in December 2022, and Google’s quality-rater guidance treats trust as the most important component.

Internal linking: comprehensive content naturally supports internal links that pass authority and guide users to related sector/service pages; Backlinko notes structured content with clear headings and semantic HTML “helps search engines surface your information more effectively. And since AI systems often pull from search results and indexed content, that structural clarity carries forward.”

How AI/LLM search changes on-page strategy (2025–2026): AI Overviews and answer engines (ChatGPT, Perplexity, Gemini) increasingly synthesize answers rather than list links. Scale check from a primary source: Pew Research Center found that ~18% of all Google searches in its March 2025 study generated an AI summary, and 58% of US adults ran at least one search that produced an AI summary (dataset of 68,879 searches), substantial and growing, but not yet a majority of searches. (Semrush’s 10M-keyword tracking shows AI Overview presence rising from 6.49% in Jan 2025 to a 24.61% July peak, settling ~15.69% by Nov 2025.) Practical playbook: keep the search-first foundations (crawlable IA, internal links, topical authority) and add answer-first formatting, short summary at the top of each page, question-style H2s, lists/bullets, named-source citations, clear authorship, because AI Overview citations skew heavily toward top-10 organic results.

4. On-Page Technical SEO

Title links, write descriptive, concise, unique <title> text; avoid vague (“Home”) or stuffed titles; Google may rewrite poor titles. Keep titles short enough to avoid truncation (aim under ~600 px). Match the page’s primary language.

Meta descriptions, there is no hard limit, but Google truncates the snippet to fit the device; write a unique, accurate description per page (industry convention ~150–160 characters) and avoid duplicates. Google may substitute its own snippet.

H1/H2 hierarchy, one descriptive H1 stating the page topic/keyphrase; question-style H2s that map to buyer sub-questions (also strong for AEO). URL slugs, short, lowercase, hyphenated, keyword-bearing (e.g., /sectors/public-sector-civic). Image alt text, descriptive, accurate; supports accessibility and multimodal AI search.

Schema markup (JSON-LD is the 2026-preferred format):

  • Organization at the site root, brand entity, logo, sameAs social profiles, contact points (especially valuable for confirming a single reputable brand and feeding AI/Knowledge Graph).
  • LocalBusiness or a more specific subtype, for a service firm without a retail storefront, ProfessionalService is the appropriate subtype, anchored to the LA address.
  • Service, nested to describe each offering (e.g., “Statewide voter-education campaigns,” “Multicultural brand strategy”).
  • BreadcrumbList, improves CTR and clarifies page hierarchy.
  • Schema.org vocabulary, https://schema.org/

Local SEO for an LA-based firm serving statewide/nationally: maintain byte-for-byte NAP (Name/Address/Phone) consistency across the website, schema, and Google Business Profile, even “Street” vs “St.” can weaken entity resolution. Display the master NAP in the footer site-wide, claim/optimize the Google Business Profile, and keep directory citations consistent. Use areaServed in schema to signal California-statewide and national reach without diluting the LA local signal.

5. Citable Market Facts (each with source + URL)

Multicultural / diverse-consumer buying power (Selig Center, University of Georgia):

McKinsey (the demand/opportunity case): “We estimate that companies filling these needs could tap into $300 billion of value annually,” and “In 2019, consumer expenditures by Black households totaled approximately $835 billion”, yet Black households were 13.4% of the population but under 10% of total spending., “A $300 billion opportunity: Serving the emerging Black American consumer,” McKinsey & Company (Aug 6, 2021), https://www.mckinsey.com/featured-insights/diversity-and-inclusion/a-300-billion-dollar-opportunity-serving-the-emerging-black-american-consumer

Scarcity / representation of Black-owned, women-owned, and minority agencies:

Multilingual / language-access requirements (the public-sector demand engine):

Consumer response to brands that reflect culture (and the cost of pulling back):

Sports / entertainment multicultural audience facts (Nielsen):

6. Competitor Website Positioning (Differentiation Map)

AgencyPositioning emphasisWhat they emphasize on-siteWhat they MISS / Sax can own
Burrell Communications (burrell.com)“Largest US Black-owned agency,” 54-year legacy, cultural authorityHeritage/legacy storytelling, blue-chip CPG clients (McDonald’s, Toyota, Coca-Cola, Microsoft), cultural-segmentation thinkingAlmost entirely brand/CPG and creative; little to no public-sector/civic, language-access, or measurable civic-impact narrative
Walton Isaacson (waltonisaacson.com)Certified minority-owned, lifestyle/experiential, “most interesting agency in the world”Experiential, multicultural + LGBTQ+, entertainment/sports lifestyleConsumer-brand lens; not positioned for government RFPs, voter education, or foundation/nonprofit work
Egami Group (egamigroup.com)Purpose-driven multicultural PR/communications”Purpose-driven work that reflects the multicultural majority,” social impact, Black-consumer PRStrong on purpose/PR but light on hard, quantified outcomes and prime-AOR public-sector campaign scale
Carol H. Williams (chwadv.com)Largest independent female-owned African American agencyBroadcast/creative, Fortune 500 multicultural campaignsTraditional ad-agency framing; not a civic/public-sector or research-first content play

Messaging white space Sax can own (and that the research validates):

  1. Public-sector / civic prime-agency-of-record credibility, the ~$16M statewide recall voter-education campaign across all 58 California counties in 10+ languages is a differentiator none of the above lead with. Tie it to the VRA Section 203 mandate (legal demand) and the WBE/MBE/SBE certifications (procurement eligibility).
  2. Measurable impact, Orbit/Ascend2 and B2B-buyer research show clarity + named proof + results beat mission language; Sax should publish hard numbers (counties reached, languages, turnout/awareness lift, dollar value) where competitors publish adjectives.
  3. A research-first content cluster (“Insights”), original analysis weaving the Selig/Census/Nielsen/Edelman/Collage facts above. This simultaneously serves E-E-A-T, earns backlinks, and is the exact “original research/analysis with clear sourcing” format AI Overviews preferentially cite, turning Sax’s data fluency into search visibility competitors lack.

Recommendations

Stage 1, Architecture & messaging foundation (weeks 1–3).

  • Lock a 6–7 item, 1–2-word primary nav with a Sectors mega-menu (Public Sector & Civic / Nonprofit & Foundations / Sports, Entertainment & Culture) and a Services mega-menu. Use mega-menus only because there are genuinely multiple sector/service pages to group (NN/g + Orbit guidance).
  • Write a descriptive H1 in plain language (“We build multicultural, multilingual campaigns that move people to act, for government, civic, and culture-defining brands”), include the target keyphrase, and pass the “backyard BBQ test.”
  • Convert “we”-led copy to “you”-led copy on the homepage and sector pages; reserve “we/our” for proof and About.
  • Benchmark to advance: homepage and three sector pages each have a single descriptive H1, a specific verb CTA above the fold, and named proof within the first two scrolls.

Stage 2, Proof & conversion (weeks 3–6).

  • Build an Impact page and per-sector case studies leading with quantified outcomes (the $16M / 58 counties / 10+ languages campaign as the flagship). Add named testimonials with photos and recognizable client logos (NFLPA, Memphis Grizzlies, Golden State Warriors, LA Opera, Universal Music Group).
  • Implement the schema stack (Organization + ProfessionalService + Service + BreadcrumbList) in JSON-LD; set footer NAP and Google Business Profile to byte-for-byte match; add areaServed for CA-statewide + national.
  • Benchmark: Rich Results Test and Schema.org validator pass with zero errors; GBP and site NAP identical.

Stage 3, Research-first content cluster & AEO (weeks 6–12, then ongoing).

  • Publish 4–6 cornerstone “Insights” pieces built on the primary facts above (e.g., “Why language access is a legal mandate, not a checkbox,” “The $300B multicultural opportunity in public-sector outreach,” “What Nielsen’s data says about multicultural sports audiences”). Format answer-first: short summary up top, question-style H2s, lists, named-source citations, visible author bios with credentials.
  • Add 3–5 internal links from each Insights piece into the relevant sector/service pages.
  • Benchmarks that would change the plan: Track AI Overview citation appearances and top-10 organic rankings for target sector keywords in Google Search Console. If AI-summary presence on your priority queries climbs toward Semrush-style highs (>20%), prioritize even tighter answer-first summaries and FAQ/Service-nested schema. If branded/sector queries are stuck below page one after 3–4 months, expand depth and earn citations rather than adding length for its own sake.

Decision triggers:

  • If a public-sector RFP pipeline is the priority → make Public Sector & Civic the first sector page, lead the homepage hero with the recall campaign, and foreground certifications.
  • If sports/entertainment growth is the priority → lead with the Nielsen multicultural-audience data and the NFLPA/Grizzlies/Warriors/UMG roster.
  • Either way, the research-first content cluster is non-optional, it is the highest-leverage move for both differentiation and AI-era discoverability.

Caveats

  • Selig Center recency: the most recent freely public Selig figures cover 2021 data ($18.5T total; $3.2T combined; $2.1T Hispanic) and 2020 for the Black-only figure ($1.6T / 9%). Newer editions exist but are sold for purchase; treat any “2024/2025” buying-power numbers from secondary outlets as projections and label them as such.
  • AEO statistics provenance: the most reliable AI-Overview scale figure is Pew’s (~18% of searches; 58% of adults). Many other AEO stats (citation rates, conversion multiples, Gartner’s “25% by 2026”) come from commercial SEO vendors (Surfer, HubSpot, CXL, Frase), directionally useful but not peer-reviewed; cite Google Search Central and Pew as the primary anchors.
  • 4A’s owner-diversity data is paywalled and self-reported (voluntary, non-random sample, no margin of error). Use the Ad Age/4A’s blog summaries directionally and pair with the more robust Census/Brookings business-ownership figures.
  • Schema/local-SEO specifics synthesize Google/Schema.org primary docs with practitioner guides; validate every implementation in Google’s Rich Results Test before launch, as eligibility and rich-result support change frequently.
  • Avoid the unverifiable claim “less than 2% of US agencies are Black-owned and woman-led.” The defensible, individually sourced substitutes are: ~3% of US businesses are Black-owned (Census/LendingTree), ad industry leadership is ~90% white (4A’s via Ad Age), and Black women own under 1% of all US businesses (Brookings).