But buyers who don't know EKU yet search in a different language. If EKU doesn't write this category's vocabulary, someone else will. Six independent evidence streams — AI association, search demand, classification, SEO authority, real search behavior, and measurement infrastructure — point to the same conclusion: EKU is findable by name, not yet by the problems it solves.
Two different, non-interchangeable brand-share reads: 99.2% is a real Search Console measurement (13 Aug 2026 clean export). A separate modeled traffic estimate in our analysis puts it at 96.6% (10 Aug 2026) — different method, same direction, not the same number.
Circle diameter is proportional to each node's own score; solid ring = Authority Score, double ring = AI Visibility (different metrics — never size-compare across ring styles) — Authority Score (SEO/backlink metric, large number) or AI Visibility where that's the only measure. Color = group: ■ direct specialists · ■ enterprise platforms · ■ OEM-native (Caterpillar) · ■ packagers · ■ scale/EaaS · ■ adjacent. EKU's circle is drawn to the same scale — that's the point. Sources: Kanbina Lumina market intelligence dataset, normalized pulls 10–17 Aug 2026 (per-node dates in the master).
The whole story in five scenes — every scene is clickable and opens the full evidence below.
AI holds EKU's complete file. Nobody asks it to open it. That is the cost this engagement is moving.
Specialists own components. Platforms own scale. OEMs own their own iron. The mixed-brand orchestration space between them is measured — and open.
"Electronics company" removed; correct power-generation identity live; new description approved.
Why: Google was teaching the market the wrong company.Next: US verification visit + services list.297 clicks · ~7K impressions · 4.2% CTR · avg pos 9.1 (14 Jul–13 Aug).
Why: without a clean sensor, no claim of progress can be proven.Next: monthly like-for-like reads.27→29 US keywords · 208→248 referring domains / 923 backlinks (12 Aug).
Why: the before/after story needs instruments turned on.Next: key events configuration (named list ready).Plant-wide control and orchestration; digital twin comparing live behavior against expected margins. Data centers, mining, behind-the-meter.
Monitoring and analytics — energy, downtime, the operation made visible. One ecosystem, never presented separately.
Phase 2 working-final delivered: messaging, channels, content engine, partner map — validation gates named.
Gate: technical sign-off on claims.Growth architecture: revenue map, 12-month roadmap, priority stack, executive handover — measured against the repaired baselines.
Gate: EKU inputs (CRM, interviews).The options, ranked by priority — run by EKU's team, a partner, or Kanbina. EKU's decision.
See the ranked options below.This is the 2026-07-31 validated starting point for everything below: Power Asset Readiness is the name for a capability EKU already has, not a category the market has granted it yet.
Control, integration, power management and lifecycle are not separate offerings — they are one world with four faces. Integrate · Automate · Optimize (below) is how that world is promised to a buyer.
Why this narrative exists
OEMs sell assets. Packagers assemble them. Controls vendors sell components. What no one in the measured cohort names is the outcome the buyer actually pays for: purchased equipment becoming operable, visible, software-ready capacity. Power Asset Readiness names that outcome — which is why it organizes everything EKU already does instead of inventing something new.
First-mover window — happening now
Yotta 2026 (Sept 28–30, Las Vegas) bills itself as the world's largest event at the intersection of AI, energy and infrastructure — 6,000+ attendees, 250+ speakers, framed around a potential $7 trillion in hyperscale data-center investment and a 70-gigawatt energy gap. Its sponsor floor is full of power players — EaaS operators, genset packagers, engine OEMs. The ones speaking there are writing the vocabulary of exactly the space EKU operates in — and none of them describe the mixed-brand readiness capability EKU has. A category that doesn't yet have a name goes to whoever names it credibly first.
The participant map is now drawn — as language intelligence, not a sponsorship case. Of the sponsor floor, at least four names already sit inside EKU's mapped orbits (an engine OEM named on EKU's own Maximus page, a rental-power packager, an enterprise power-management platform, and a Rehlko company from the authority cohort), with five more packager/EaaS players on the watchlist. The vocabulary that floor speaks — bridge power, commissioning power, behind-the-meter, AI factories, the energy gap — is the vocabulary EKU's data-center content must answer in.
Full participant-by-participant map in the working workbook.
yotta-event.com, reviewed 14 Aug 2026 · participant orbit-mapping: Working Hypothesis, Edward validation pending.
Defensive domains — cheap insurance on the category name
If Power Asset Readiness works, someone else will try to own its address.The positioning decision
| Job in the system | Approved working language |
|---|---|
| Commercial narrative | Power Asset Readiness — the organizing story: turning purchased assets into operable, visible, software-ready capacity. Not an SEO keyword; the narrative. |
| Buyer descriptor | Controls and monitoring for mixed-brand generator fleets — the operating environment in plain buyer language. Short form: mixed-brand fleets. |
| Technical proof | Chip-to-Cloud architecture and cross-manufacturer integration — the engineering depth that convinces the technical role. |
| Search-demand entry language | Generator controls · remote monitoring · idle management · fleet monitoring · power generation controls — the terms with existing demand that bring buyers into the story. |
Approved in session as working narrative (Jul 31, 2026). Note: no meaningful search signal was found for "multi-OEM power systems" in the analyzed data — that phrase does not lead acquisition. Full rationale in D04.
"OEMs provide the assets. Packagers assemble the package. EKU provides the layer of control, integration, software and lifecycle that turns that equipment into energy capacity that is operational, visible, and manageable."
industrial power controls · generator control systems · genset integration · generator automation · remote monitoring · fleet management · idle management
"Equipment does not automatically equal operational capacity." Buying and assembling generation equipment does not guarantee the asset is ready to run.
Organizes GeneSys, Maximus, Sophia, ESC and lifecycle support under one idea instead of a disconnected product list.
In plain terms: companies buy expensive generation equipment from different brands. Owning it is not the same as being able to run it — connected, monitored, coordinated, ready. Power Asset Readiness names the closing of that gap: making purchased equipment ready to operate as one system. It is the umbrella over everything EKU sells, not another product beside them.
Why "OEM-agnostic" was retired — three reasons:
The trade-off, stated openly (keep vs. retire):
| Keeping "OEM-agnostic" | Using "mixed-brand generator fleets" | |
|---|---|---|
| Strength | Familiar internally; short; investor audiences know it | Matches measured buyer language; claim-safe; approved |
| Weakness | 0 search demand; implies universal compatibility | Longer; needs internal adoption effort |
| Opportunity | — | Own an unclaimed, searchable phrase in the market |
| Threat | A single incompatibility becomes a credibility event | None measured — the phrase describes operating reality |
Where each lane lives: buyer-facing material uses mixed-brand generator fleets; investor-facing decks retain the legacy term by EKU leadership preference. Both lanes are legitimate — they never mix.
Decision validated in session, 31 Jul 2026 · demand evidence: Kanbina Lumina market intelligence dataset, Jul–Aug 2026
Named: 503/744. Nobody typed the name: 0/5,176.
Our AI-answer testing, Jul 2026
Idle management: live search equity. Generator paralleling controller: low-difficulty, real demand.
Search Console + our keyword analysis, Jul 2026
Prior GBP category ("Electronics company") fed VFD confusion. New proposed category/description below.
GBP audit, Jul 2026
Authority Score 18 vs. 29–37 cohort — an industry SEO/backlink metric, not overall brand authority.
our analysis, 10–12 Aug 2026
99.2% of GSC clicks are on EKU's own name; 0.8% non-branded (full clean export, ~9-month window). A separate modeled traffic estimate puts brand share at 96.6% — a different method, same direction.
Google Search Console, clean export, 13 Aug 2026
Zero key events configured in GA4 across ~115 sessions this period, despite real engagement (5 form starts, 3 downloads, a referral from bauma.de â the world’s largest construction & mining machinery trade fair).
GA4, 14 Jul–12 Aug 2026
Proven capability. A market that hasn't found it yet.
Click any vertex above — each proof opens its own deep dive here
Six independent measurements, one conclusion. Presented internally as a working category; presented to EKU as a strategic framework; never yet as an established market category that replaces existing keywords.
Google Business Profile — the proposed re-description, and what was already fixed
The profile categories are already changed — the "Electronics company" classification that fed the VFD confusion is gone (GBP audit fix, executed).
Name stays EKU Power Drives — real-world name, no keyword stuffing, which protects the listing.
Pending change: the profile description. Proposed draft: "EKU Power Drives turns purchased power generation equipment into systems operators can run, see and control. We integrate, automate and optimize mixed-brand generator fleets — controls, monitoring and power management, from chip to cloud — for data centers, mining, oil & gas and industrial sites. Products include GeneSys controls, the Maximus orchestration platform and Sophia monitoring."
Execution status (17 Aug): the Germany profile can take the new description as soon as it is approved; the US profile is gated on account verification (in-person verification visit being scheduled). NAP unification is deliberately ON HOLD: a US relocation begins partially in November and the Woodlands location closes by February — unifying addresses now would mean doing the work twice. Both addresses stay listed until the move plan is confirmed.
Maximus operates as the upper orchestration layer. Sophia executes analytical monitoring. Integration proven via Chip-to-Cloud technical trial.
EKU already ranks #1 for its own branded query and is cited in a Google AI Overview describing Maximus as an orchestration platform for power generation. Separately, "Maximus controller" ranks #18 and "Sophia integration" registers real volume. The commercial opportunity is connecting that existing footprint to a proof-led narrative that survives contact with a skeptical buyer.
The buyer's real choice
Caterpillar is the alternative — and that is EKU's best argument.One fact makes this sharp: EKU's own Maximus page names Caterpillar as a compatibility example, while Caterpillar sells its own control layers as the alternative in the same buying decision. That tension is the sales argument — independence versus lock-in — and it is why Caterpillar sits in the universe map above.
* Caterpillar shown as the worked example — EKU's own live Maximus page already names it as a technical-compatibility example. The same analysis extends to the rest of the measured cohort as benchmark pulls land.
cat.com's digital gravity is brand and ecosystem — 45% direct traffic, a referral network of its own dealer and financing sites — while its top non-branded search terms are construction equipment, not power generation. In the measured keyword gap it holds only weak positions on generic power terms ("genset" #65, "power plant" #45).
of cat.com traffic is direct — brand gravity, not content
of its organic entry is non-branded category language — even giants live on category terms
cat.com's position for "genset" — the language is takeable
of its traffic already arrives from AI assistants (~56K visits/quarter) — AI answerability is countable, not theoretical
EKU live Maximus page, 13 Aug 2026 · competitive scan of public Caterpillar product information, 13 Aug 2026 — EKU engineering validation required before external use · cat.com digital benchmark: external traffic-intelligence dataset, Mar–May 2026 data, compiled Jul 2026 · keyword positions: our 30,000-row keyword-gap analysis, 31 Jul 2026. Caterpillar's Authority/AI-Visibility scores arrive with the next benchmark pull.
That layer is what EKU sells. Hypothesis carried by EKU marketing + Kanbina, every claim gated behind EKU technical sign-off — this engagement is structured to validate it before anyone says it externally.
Open questions, not resolved claims
Self-cannibalization: is there a reason Maximus and Sophia would cannibalize each other — or cannibalize EKU's own hardware/GeneSys sales — by making it easier to run someone else's equipment well? Not answered here; worth asking directly rather than assuming the answer is no.
"Behind-the-Meter" has no visible page owner. The term shows up in buyer-language research but isn't claimed by Maximus, Sophia, the microgrid page, or any other page in the current site architecture — an orphaned term, not yet assigned in the keyword-ownership table below.
The changes, side by side
Every recommendation is checkable against what is already written, already searched, or already said in a working session — not invented from scratch.
"Multi-OEM Compatible Platform for Modern Power Systems Integration"
Leads with a technical category.
"We turn the generation equipment you already bought into a system you can operate, see and control."
Leads with the outcome the buyer wants.
Why: "multi-OEM" and "mixed-OEM" have no search demand — and no competitor uses them. The buyer isn't searching for a platform; they want to operate what they already have. For what: the headline is the first filter every buyer applies; leading with the outcome puts EKU in the consideration set before any technical comparison starts. Alternate phrasings are held in the D05 working file for the final selection session.
"Power generation is becoming more complex. Operating it shouldn't be."
EKU's best message lives three clicks deep.
"Power generation is becoming more complex. Operating it shouldn't be."
Moves to the homepage and heads Control & Automation.
Why: it's the best buyer language in all of EKU's material, and it describes exactly the open space. It's already written — it just needs to move up front.
Decarbonization · net-zero · sustainability · "low-carbon infrastructure"
Sustainability as the main argument.
Equipment availability · downtime · cost to operate · ready-to-operate system
Sustainability becomes a benefit, not the headline.
Why: sustainability is the ground ComAp already occupies, and it's not the buying criterion in mining, oil & gas or construction. There it's decided on availability and cost.
Confirmed
The branded query "ekupd.com Maximus" returns EKU's Maximus page first, cited in a Google AI Overview as an orchestration platform for power generation. Proves indexation and message-ingestion — not non-branded discovery; the query is navigational.
Evidence Register E29, 13 Aug 2026
Confirmed
"Maximus controller" — 40 searches/mo (US), EKU ranks #18. "Sophia integration" and "Maximus controls" also register.
Kanbina intelligence dataset, 10 Aug 2026
Needs Validation
A "100+ installations" figure appears in EKU-sourced material. First-party claim, not independently verified — held out of all client-facing copy until a technical owner confirms sourcing.
Internal claim ledger, Aug 2026
External Research
Direct review of ekupd.com/maximus (13 Aug 2026): the live page currently uses "OEM-agnostic by design" as a headline claim, naming Caterpillar, MTU and Waukesha–INNIO as technical compatibility examples — consistent with this deck's claim-safety rule that named OEMs are described only as technical compatibility, never partnership or endorsement. Worth flagging directly rather than smoothing over: this is the same "OEM-agnostic" phrasing Edward pushed back on in the 24 Jul 2026 working session (he prefers "multi-OEM"), so the live site and the internally-preferred language aren't yet aligned. The page's four named industry applications include data centers — one of EKU's own four stated verticals, not a Kanbina addition.
Live page review, ekupd.com/maximus, 13 Aug 2026
Founder/Interview Insight · 24 Jul 2026
Edward's own refinement of the offer: technology services, data services and lifecycle services as the three core categories — plus two optional, not-currently-pursued revenue streams (engineering consulting and packaging). Feeds the D05 messaging structure; not yet tested as external page architecture.
EKU working session notes, 24 Jul 2026
Where this comes from: our keyword-demand analysis (13 Aug dataset) — every row is measured monthly search volume in the US, screened for fit with the ecosystem's territory. None of these are current rankings. They're candidate demand anchors pulled from the raw keyword strategy pool, flagged "OPEN" pending SERP-fit and technical validation — shown here as the scale of the opportunity, not a promise of where EKU ranks today. Reading the volumes: each number is measured US searches per month for that exact term; the industry benchmark that says whether 320/mo is large for this category arrives with the next cohort pull — flagged, not guessed.
| Candidate term | US vol./mo | Difficulty | Intent | Candidate page |
|---|---|---|---|---|
| power management system | 320 | 27 | Informational | /genesys-switchgear or new PMS page |
| energy management control system | 320 | 19 | Informational | /maximus |
| microgrid control system | 210 | 18 | Informational | New cluster page / Maximus |
| power generation software | 210 | 48 | Informational | /maximus |
| energy management platform | 170 | 36 | Informational | /maximus |
| energy grid software | 140 | 47 | Informational | /maximus |
| energy grid management software | 110 | 23 | Informational | /maximus |
| industrial energy management software | 110 | 13 | Commercial | /maximus |
| energy management software platform | 70 | 22 | Informational + Commercial | /maximus |
| power generation orchestration platform (category term) | no reported volume in this pull | — | Narrative | /maximus |
Command Center, sheet 09 Maximus Keywords, 13 Aug 2026. All rows status "OPEN" — candidates require SERP-fit and technical-fit validation before becoming targets. D04D05D07
Kanbina's own synthesis, built to stop every keyword from getting force-fit onto the Maximus page. Each page earns a distinct territory instead of competing with its own family of pages.
| Page | Candidate territory |
|---|---|
| Maximus | Power management, system-level orchestration, distributed-energy coordination, load management, chip-to-cloud narrative |
| Sophia / monitoring | Generator monitoring, wireless monitoring, fleet telemetry, remote visibility, predictive-maintenance data |
| Microgrid solution page | Microgrid controller, microgrid management, behind-the-meter coordination |
| Smart PDU page | Smart PDU and power-distribution terms |
| Controls / GeneSys page | Generator control panel, genset controller, paralleling, synchronization, ATS/controller terms |
| Learning Center | Power Asset Readiness, benefits/comparisons, market education, category definitions |
| Segment pages | Data centers/hyperscale, mining, oil and gas, rental/temporary power — only after segment validation |
Kanbina interim executive analysis, 13 Aug 2026. Guardrail carried forward: competitor-brand terms should only appear in compatibility/comparison content once technical compatibility and claim permissions are confirmed.
Data centers is one of the four industry applications EKU's own live Maximus page already names — not a segment Kanbina is proposing from nothing. External market evidence is unusually strong and well-dated: US data center power demand is projected to roughly double, ~31GW to ~66GW, from 2025 to 2027; PJM and Dominion interconnection queues each carry roughly 70,000MW of pending requests; equipment lead times run 128–144 weeks for transformers and 40–60 weeks for switchgear. A 16-role buyer-committee map and a 15-dimension readiness scorecard exist internally (Kanbina IP, not reproduced here).
And the search battlefield is measured — every data-center power term, who holds it today:
| Term | US vol./mo | EKU | Vertiv | Schneider | Caterpillar | Generac |
|---|---|---|---|---|---|---|
| data center power solutions | 1,300 | — | #44 | #22 | #52 | #36 |
| data center backup | 390 | — | #47 | #37 | #18 | #13 |
| generator paralleling switchgear | 110 | — | #45 | #3 | #11 | #25 |
| data center backup power systems | 50 | — | #71 | #43 | #22 | #1 |
EKU: zero presence on every measured term. The vertical EKU's own page names is, in search, entirely owned by others today — the entry cost and the size of the prize in one table. Named companies here are search-results competitors only; no commercial relationship implied.
Status: Working Hypothesis — Active Validation Priority, not Confirmed as an EKU-served segment. Guardrails carried forward: never claim to "solve" PJM/ERCOT interconnection queues, never convert market-size percentages into EKU revenue-capture claims, never claim to eliminate UL 9540A testing requirements or guarantee Tier IV uptime.
Data Center Buyer-Fit Evidence Sprint (external, sourced, dated register), compiled 13 Aug 2026.
Every figure below carries its own source and date. None of them replace or average against the others — the conclusion leads, the number backs it up.
The real, measured number: 99.2% of Search Console clicks are on EKU's own name, 0.8% non-branded (full clean export, ~9-month window, 13 Aug 2026). A separate, differently-scoped modeled estimate — 96.6% of inspectable traffic from three brand queries — points the same direction using a different method; the two are not the same measurement and shouldn't be quoted as if they were.
GSC clean export, 13 Aug 2026 · modeled traffic estimate, our analysis, 10 Aug 2026
Prompted with the EKU name, AI tools associate it correctly 503 times in 744 test responses. Asked the underlying buyer problem with no brand named, EKU appears 0 times in 5,176 responses.
our AI-answer testing, Jul 2026
Careers, Leadership, About and Contact out-rank product and solution pages for real clicks. The site is currently better optimized to be a company than to be a supplier.
GSC clean export, 13 Aug 2026
"Idle management" already gets 568 monthly impressions at position 4–6 (page one) in EKU's own Search Console — and converts to roughly one click. Real page-one visibility, almost no capture. "Idle management system" separately carries real, if modest, measured monthly search volume. Together they're the clearest, lowest-risk first move in the entire keyword set.
Google Search Console, presented live 24 Jul 2026 · our keyword analysis (20/mo, Tier 1), 10 Aug 2026
EKU's Authority Score — an industry SEO/backlink-strength metric, not a measure of overall brand or commercial authority — sits at 18. The focused comparison cohort (DEIF, ComAp, Deep Sea, Basler) runs 29–37. 248 referring domains / 923 backlinks (12 Aug) vs. hundreds to thousands of referring domains for the cohort.
Kanbina intelligence dataset, 10 Aug 2026 · Evidence Register E04, 12 Aug 2026
US organic keyword coverage nearly doubled year over year (Jul 2025 to Jul 2026). Real, positive, and still small enough that one good quarter of work can move it meaningfully.
domain-rank history, our analysis, Aug 2026
GA4 shows zero configured key events across the entire measured window — even though real signals already exist (5 form starts, 3 file downloads, a real trade-fair referral from bauma.de (bauma: the world’s largest construction & mining machinery fair)). The instrumentation gap, not a traffic gap, is what's blocking a credible before/after story.
GA4, Jan–Aug 2026
A second, separate AI metric in our analysis alongside AI association: AI Visibility measures share of voice inside AI-generated answers. EKU 14, Deep Sea 27, DEIF 28, ComAp 25, VoltaGrid 18, Basler 16, PowerSecure 23, Mesa Power 25, Caterpillar 43, Vertiv 67, Schneider Electric 85 (normalized pulls, 12–17 Aug — per-domain dates in the master). AI Audience (a related but distinct volume metric) shows an even starker gap: EKU 0, Deep Sea 3.5M.
Kanbina AI-visibility dataset, 12–17 Aug 2026
This is the journey the evidence supports — where buyers start, what they type, and where EKU appears or doesn't. It reframes the landscape work (D01) and drives the messaging system (D05).
New site, mixed fleet, downtime event, expansion, energy-cost pressure. No vendor names yet — only a problem.
EKU visibility today: none — this stage happens away from any channel EKU owns.
"Generator control panel," "remote generator monitoring," "idle management" — and increasingly they ask an AI assistant the same question.
EKU: 0 of 5,176 problem-based AI answers; largely absent from search page one.
Specialists (DEIF, ComAp, Deep Sea) and the OEM-native option (Caterpillar's control layers) dominate what's findable.
EKU enters only if someone in the room already knows the name.
Compatibility, protocols, references, install base. This is where Chip-to-Cloud and mixed-brand integration win — if they're visible and claim-safe.
EKU's proof story exists internally; the public version is still thin.
Seven roles weigh in (below). Post-sale, lifecycle support decides whether the account grows.
EKU's strongest stage — but buyers can't reach it if they never reach EKU.
Journey stages: Kanbina synthesis of the sourced evidence in this deck (search data, AI-answer testing, GSC behavior, working-session input). Stage-by-stage validation with real EKU buyers: pending — flagged, not assumed.
Each role enters through a different door and reads different proof. Positioning has to survive all seven, not just the technical buyer.
Buyer Committee Swimlane (AV06), supports D02 + D04. Data Center Operator carries elevated Needs Validation status — segment not yet confirmed with EKU. HR added per Founder/Interview Insight: highly specialized controls engineering talent is scarce and expensive to concentrate on a single internal team — EKU's packaged intelligence is also a talent-cost argument, not only a technical one. CFO angle confirmed directly by Edward (24 Jul 2026): better energy management reduces the amount of new generation equipment a buyer needs to purchase in the first place — a capex-avoidance argument, not only an opex-savings one.
The seven roles above describe who sits in the room; this lens describes which organization convenes the room. Each profile is built from the evidence already in this document and is explicitly labeled where direct validation is still pending.
Who decides: the electrical engineering lead specifies; procurement executes under project deadlines and pre-qualified vendor lists.
Research behavior: spec-driven — component-level search language ("generator control panel," switchgear and protocol specs), manufacturer documentation, and vendors already proven on prior projects.
Prioritized entry point: technical documentation depth + the integrated-controller compatibility list (D05 claim set, pending EKU engineering sign-off) — spec-ready content wins this door.
Status: Working Hypothesis — built from measured search language and journey evidence above; direct EPC interviews pending (Needs Validation).
Who decides: the VP of energy/infrastructure with an owner's engineer; timelines are driven by time-to-power.
Research behavior: outcome language rather than component language — readiness, uptime, behind-the-meter capacity — and increasingly AI-assisted discovery (the 0-of-5,176 finding above is sharpest here).
Prioritized entry point: the Maximus orchestration story + Power Asset Readiness narrative, carried into industry events (Yotta-type rooms) where this segment actually gathers.
Status: Working Hypothesis — consistent with the data-center keyword cluster and event-language mapping (AV09); Data Center Operator segment itself carries elevated Needs Validation status.
Who decides: standards and engineering committees; purchasing runs through qualified-vendor lists and formal RFPs — the longest cycle of the three.
Research behavior: procurement platforms and directories, consultant-written specifications, compliance documentation — less open-web search than the other two segments.
Prioritized entry point: interoperability and compliance proof, plus presence on the procurement-platform surface (mapping pending, per the discovery-surface note above).
Status: Working Hypothesis — lowest direct evidence of the three segments today (Needs Validation); no utility-side interviews conducted yet.
Segment profiles: Kanbina synthesis of sourced evidence in this document (search-language analysis, journey mapping, channel findings). Each profile upgrades from hypothesis to validated map through the interview program already flagged in D02. D02
Real, sourced demand — not a hypothetical gap. Every row below is a term with measured US monthly volume.
| Buyer search language | Cluster | US vol./mo | Difficulty | EKU rank today |
|---|---|---|---|---|
| smart PDU | Energy management & distribution | 480 | 7 (low) | — |
| generator monitoring system | Data intelligence & remote ops | 320 | 2 (low) | — |
| generator control panel | Controls & automation | 260 | 2 (low) | — |
| microgrid controller | Energy management & distribution | 260 | 28 (moderate) | — |
| power plant control system | Controls & automation | 170 | 3 (low) | — |
| smart power distribution unit | Energy management & distribution | 170 | 8 (low) | #28 |
| remote generator monitoring | Data intelligence & remote ops | 140 | 0 (open) | — |
| black start generator | Blackstart & auxiliary energy | 70 | 14 (low-mod) | — |
| Maximus controller (branded) | Product | 40 | 1 (open) | #18 |
| idle management system | Fluid management & efficiency | 20 | 0 (open) | — |
| Sophia integration (branded) | Product | 20 | 0 (open) | — |
Kanbina keyword-demand analysis, 10 Aug 2026. Volume/difficulty are directional modeled estimates, not revenue forecasts. D05D07
Benchmark — who already owns this language
From our 31 Jul keyword-gap analysis (EKU vs. Deep Sea, ComAp, DEIF): the cohort ranks and EKU is absent for "ats controllers" (140/mo), "generator paralleling controller" (70/mo), "diesel engine controller" (50/mo), "controller gen" (40/mo). At domain level: EKU 19 organic keywords vs. Deep Sea 337, ComAp 395, DEIF 439.
Honest limit: a term-by-term competitor-rank column for the full table above requires a fresh benchmark pull, which is scheduled as the next benchmark pull. Marked pending rather than estimated — no invented ranks.
Kanbina keyword-gap analysis, US Desktop, 31 Jul 2026 · benchmark refresh scheduled, Aug 2026
Bring assets from different vendors and eras onto one control and monitoring layer.
Turn manual, site-by-site operation into coordinated, rules-based control.
Use the visibility that integration and automation create to run assets closer to their real capacity.
"We are now separating commercial competitors, digital competitors and ecosystem-scale benchmarks so we can see not only who competes with EKU, but who is actually capturing buyer attention online." Every node in the hero orbit reappears here in its tier, with the deep-dive behind it. Four real tiers, not one flat list of rivals. Every set, its "why include" reasoning, and its validation status come directly from the working competitor register — internal hypotheses, not settled facts, until Edward validates each one.
EKU wins its own name outright. Not a contested arena.
Direct-controls cohort (DEIF, ComAp, Deep Sea) holds this arena today; EKU is largely absent from page one.
EKU appears in 11 tracked AI Overview keyword slots — a real, measured toehold, small relative to the cohort.
248 referring domains (12 Aug) vs. hundreds (PowerSecure, VoltaGrid) to well over a thousand (DEIF, ComAp) for the cohort.
PowerSecure and VoltaGrid occupy adjacent integrator and market-signal positions rather than head-on control-panel competition. New measured signal (17 Aug): PowerSecure runs an active paid-search program on exactly the data-center power language â "data center microgrid" (#1 paid), "prime power diesel generator" â live proof that the demand EKU’s vertical hypothesis names is worth buying ads for.
Two sources disagree and haven't been checked against each other. Reading A: EKU's own Vulcan Heating Systems page (ekupd.com/vulcan-heating-systems) shows a recurring pattern of high impressions and near-zero clicks — real, owned content, underperforming. Reading B: the 24 Jul 2026 working-session notes describe "Vulcan" as an unrelated UK brand creating keyword/search-data confusion. Held open rather than silently resolved either way — the two readings point to different fixes (own-content optimization vs. keyword-collision cleanup), so the distinction matters before either is acted on.
The white space, as a bubble map — bubble size = organic keyword count
Concrete missing-keyword examples EKU doesn't rank for at all today, that its direct-specialist competitors do: "ats controllers" (140/mo), "generator paralleling controller" (70/mo), "diesel engine controller" (50/mo), "can you play ats with a controller" (70/mo), "controller gen" (40/mo).
Kanbina keyword-gap analysis, ekupd.com vs. deepseaelectronics.com / comap-control.com / deif.com, US Desktop, 31 Jul 2026.
The white-space matrix — the pre-engagement map, now drawn with real data
The visible, mixed-brand, system-level position is empty. That emptiness is the strategy.Read it in one line: the specialists are visible but component-anchored; the giants and VoltaGrid are visible but brand-locked or asset-owning; EKU already sits in the right column — the work is the vertical climb, not a repositioning. This is the same matrix logic used in the original proposal, now placed with measured data instead of estimates.
Positions: horizontal from product architecture (component vs. system scope, our competitor register); vertical from Authority Score + AI Visibility; bubble size from organic keyword counts, 31 Jul–14 Aug 2026 datasets. Caterpillar is placed with its measured 17 Aug benchmark (AS 61 / AI-Vis 43, ecosystem scale); Schneider directional pending its pull.
Thread to EKU: they already own the words EKU’s buyers type.
Thread to EKU: they set the language ceiling EKU’s claims must stay credible under.
Thread to EKU: they decide who gets cited — the shortest route to AI-answer visibility.
Thread to EKU: they prove the demand — EKU serves it for everyone who can’t buy their scale.
AS = Authority Score (industry SEO/backlink metric). AI = our AI Visibility measurement. Different metrics, shown separately, never combined into one number.
Strong overlap in genset, microgrid and brand-agnostic control language.
Owner question: where does EKU win or lose vs. ComAp in real projects?
Power management, genset, battery and solar controller coverage.
Owner question: is DEIF a direct alternative, a component, or both?
Controller and load-sharing visibility in generator control markets.
Owner question: does EKU compete at controller level or system level?
Relevant to generator protection, excitation and control terms.
Owner question: where is Basler complementary vs. substitutable?
Generator power management, synchronizer and industrial controls presence.
Owner question: which Woodward offers appear in EKU deals?
Relevant to protection, automation and mission-critical power controls.
Owner question: which buyer segments overlap with EKU?
Not a specialist peer — the OEM whose own control layers are the default alternative in any fleet that already owns Cat equipment. The decisive tension (compatibility example vs. closest alternative) is unpacked in the Maximus section above.
Owner question: in real deals, how often is "just use Cat's controls" the option EKU displaces?
All Set A entries: status NEEDS TECHNICAL VALIDATION. Command Center, sheet 06, 13 Aug 2026 · AI Visibility, 12–17 Aug 2026 · Caterpillar normalized benchmark: AS 61 / AI-Vis 43 / ~194K US organic keywords — ecosystem scale, never compared like-for-like with specialists.
Large visibility benchmark for energy management and microgrid control.
Owner question: which claims are credible for EKU at smaller scale?
Narrative benchmark for automation, control and digitalization.
Owner question: which category language overlaps without overreaching?
Power generation and plant software benchmark.
Owner question: is comparison relevant to EKU's target project size?
Benchmark for control-to-analytics and operational performance narrative.
Owner question: which proof points would EKU need to compare credibly?
Comparison scoped to exact product URLs only, never full-domain — comparing EKU to a Schneider or Siemens root domain would be misleading at this scale.
Often ranks and is cited for asset performance management language.
Ranks for power generation asset management and industry guides.
High-trust citation source for power, reliability, microgrids and VPPs.
High-trust language around reliability and resource adequacy.
Trusted research and asset-management content for power systems.
Energy-as-a-service operator, not a controls vendor. $1B Blackstone/Halliburton equity, $5.0B financing package, 1.5GW INNIO order, ABB/Halliburton/Oracle partnerships. Kanbina's own working-session read (24 Jul 2026) treats it as a marketing-execution model to study, not a like-for-like product competitor.
Packager operating in the same sub-5MW kW tier EKU competes in — a dual competitor/customer relationship, not a clean rival.
Open item: whether Mesa acts as a channel or a competitor decides which play applies — flagged for commercial validation, because the answer changes the outreach, not the data.
Major data-center-power brand; highest AI Visibility score of any company in this competitor set after Schneider.
Owner question: does Vertiv's data-center footprint overlap with EKU's stated data-center vertical?
Sits at EKU's own AI-Visibility level — a same-size adjacent player worth tracking as a proxy for how fast this tier can move.
Owner question: does RPS overlap with EKU on any real deal, or only in search data?
VoltaGrid: site crawl + press-release headlines, 13 Aug 2026 — and the tier's case study: it shows what the market pays for orchestrated behind-the-meter capacity when capital is unlimited, which is exactly the demand EKU's intelligence layer serves for everyone else (the "Desired Hypothesis" in the Maximus section). Watchlist of possible additional scale signals, from the Yotta 2026 sponsor floor: Aggreko, Mainspring Energy, Enchanted Rock, Life Cycle Power, IGSA Power — candidates only, not yet assessed. Mesa Power: named in the 24 Jul 2026 working session; measured 17 Aug (AS 28 / AI-Vis 25). AI Visibility figures: normalized pulls, 12–17 Aug 2026.
What happens with the backlink data
The cohort's link profiles are a shopping list of sources that already trust companies like EKU.Seven moves, already sequenced: freeze EKU's referring domains (248, with 923 backlinks, 12 Aug) as the dated Day-0 baseline · qualify them by type instead of counting them · mine where DEIF, ComAp and Deep Sea earn their links (industry bodies, trade press, distributor pages) · flag domains linking to two-plus cohort members but not EKU as priority outreach · target the Set C citation authorities as the shortest route into AI answers · never auto-disavow · and make every link-earning asset carry the Power Asset Readiness language, so authority accrues to the category, not just the domain.
Full plan, tab 05 of the working workbook · feeds D03, D07, D08.
This is the cost named in the opening: buyers who don't know EKU yet are already searching in a different language. The 14 → 29 keyword-coverage finding above shows movement already happening since this engagement started — the work below is how that movement compounds instead of stalling.
Site Audit errors flagged
our analysis, 10 Aug 2026
Site Audit warnings
our analysis, 10 Aug 2026
AI Overview keyword appearances
our analysis, 10 Aug 2026
GA4 key events configured
GA4, Jan–Aug 2026
Reversal path, in order: fix schema/metadata debt → build the buyer-language content layer (Layer 1, above) → close the measurement gap so the next 90 days can be proven, not just felt. Procurement directories and platforms are a third discovery surface in this workstream, alongside search and AI answers — platform mapping pending.
Each line leads with what the deliverable already says — then its state. In process is a status, not an apology: this is what a 90-day evidence engine looks like mid-flight.
What we investigated: search intent, buyer questions, competitor copy patterns, AI prompt language — the analytics tooling is one input among several — the analysis and orchestration are Kanbina's.
How it changes the deliverable: turns technical capabilities into buyer questions, outcomes and objection responses, instead of engineering-led copy.
Client-facing output: messaging house, page copy blocks, objection bank.
Guardrail: do not auto-generate unverified performance claims — every claim still needs a technical-owner sign-off.
What we investigated: traffic channels, referral domains, page performance, search clusters, landing pages.
How it changes the deliverable: prioritizes where EKU needs presence and how the site should route buyer journeys.
Client-facing output: channel role map + site/page architecture.
Still needed: LinkedIn analytics, CRM source data, partner interviews — digital-channel data is strongest here, not field sales or events.
What we investigated: keyword universe, prompt universe, citation sources, page ownership, topic research — the strongest evidence base of any Phase 2 deliverable right now.
How it changes the deliverable: creates pillar/subpage architecture, metadata priorities and AI answer targets.
Client-facing output: AI/search framework + prioritized content briefs.
Guardrail: mass content production and technical build stay out of scope.
The anchor fact: roughly 70% of revenue flows through packagers (stakeholder interview, 24 Jul 2026 — the strongest channel evidence in the engagement). Packagers and end users are the actual sales channels — and the ones carrying the pain this deliverable needs to capture. Backlinks and referring domains are a proxy signal for who's already talking about EKU, not the channel map itself.
What we investigated: referring domains, backlink gaps, co-citations, competitor partner pages — a starting signal, not the finished picture.
How it changes the deliverable: surfaces ecosystem candidates and validation questions; the real work is mapping packager and end-user pain directly, not inferring it from SEO signals.
Client-facing output: partner typology + prioritization + outreach framework, organized around packager and end-user needs.
Guardrail: a backlink or co-mention is not a commercial relationship — real partner history and direct packager/end-user input still required from EKU.
Open channel decision: should packagers become a visible public segment, be addressed with softer public language ("system integrators," "power generation packaging partners"), or stay internal-only? Working recommendation: soft public language until channel sensitivity is confirmed — packager relationships may be confidential, and public packager messaging could create channel conflict.
Command Center, sheet 02 SOW Deliverable Map + Interim Executive Review deliverable-readiness table, 13 Aug 2026 · D04 approval per Kanbina Lumina, 13 Aug 2026 (not yet reflected in a written EKU sign-off record). All twelve SOW deliverables shown; status reflects evidence readiness as of 17 Aug (Day 48), not final delivery.
| Front | Before (engagement start) | Today (measured) |
|---|---|---|
| GBP — Google Business Profile | Category "Electronics company" — taught Google the wrong identity and fed VFD confusion | Category corrected; full re-description proposal drafted and ready for approval; Spring/Woodlands address unification identified |
| GSC — Search Console | Unusable exports — no clean first-party search baseline existed | Clean, usable baseline: 297 clicks · ~7K impressions · 4.2% CTR · avg. position 9.1 (14 Jul–13 Aug window) |
| GA4 — Analytics | Property/stream status unclear; no reliable collection | Stream configured and connected, data flowing through the dashboard; key-event configuration is the named next step |
| Rank measurement | A fresh, unstable tracking setup — no defensible trend possible | Stable 120-keyword campaign confirmed (Spring, TX), with a governed correction log preventing false trends |
| Organic footprint (US) | 27 keywords · 208 referring domains · 902 backlinks (mid-Jul) | 29 keywords · 248 referring domains · 923 backlinks (12 Aug) — modest, real, and honestly labeled as not-yet-commercial-proof |
| Site technical debt | Invalid structured data + 332 non-descriptive anchor links, unquantified | Fully quantified and sequenced into the fix roadmap — coverage exists; validity is the issue being corrected |
Kanbina Lumina market intelligence dataset — normalized current layer, reconciled 14–16 Aug 2026. Each pair is like-for-like (same module, geography and scope); improvements credited to the joint EKU (Amanda) + Kanbina working team.
Every artifact here goes beyond the contracted scope. It is free for EKU by design: it is how Kanbina holds its own quality bar, and how a long-term growth partnership behaves. Scope stays protected — defined, never silently expanded.
Added value · Domain protection shortlist
Block 1 · The category set — Power Asset Readiness
The canonical home of the category — the Learning Center lives here or redirects to it. Whoever owns this owns the address of the idea.
The AI-era variant. AI assistants increasingly cite .ai domains for category definitions — cheap insurance on the new front door.
The hyphenated capture. Blocks the most common typo-registration play a squatter would try first.
Blocker — never used, never available to someone else. Costs less per year than an hour of engineering.
Blocker — protects the category from a fake "institute" or "association" wrapper built on top of EKU's idea.
Block 2 · Brand protection — the company name
The full company name. If it resolves to a stranger, every search for the full name leaks trust. Verify ownership; register if free.
Blockers around the existing primary domain — the cheapest insurance against copycat or phishing use of the short brand.
Brand + AI-era pairing — same logic as the category .ai, applied to the company name itself.
Block 3 · Product-narrative candidates — verify before buying
Matches the "Maximus by EKU" phrasing already in use. Generic "maximus" domains are unobtainable; the by-EKU form is the ownable one. Candidate — availability to verify.
Same logic for the software layer — protects the ecosystem naming pattern before it's public. Candidate — availability to verify.
The approved buyer descriptor as an address — usable for a future campaign or Learning Center redirect. Candidate — availability to verify.
What the whole list avoids: a competitor or squatter registering the category's name after EKU has spent months making it valuable — then outranking EKU for its own idea, or selling the domain back at ransom pricing.
Recommendation raised in the 24 Jul 2026 working session; list refined against the approved narrative. Availability and exact pricing: confirmed at purchase. Spend requires EKU authorization — this is the case for making that decision quickly.
Active: AV01 · AV02 · AV03 · AV07 · AV09 · AV10. Delivered: AV05. AV04 is absorbed into D08's channel-visibility decision. AV06 stays gated — defined, not silently expanded into unlimited execution.
Who runs each option is EKU's decision: your own team, a third party, or Kanbina — the ranking below orders the gaps by how much commercial ground each one closes, and each card states why it exists and what it unlocks. Nothing here is inside the current 90-day scope.
Trigger: 96%+ brand-heavy traffic + confirmed buyer-language whitespace. Turns D05–D07 strategy into shipped pages, technical fixes, and a monthly readout.
Priority 2 · Future Optional Program · 6-month operating programTrigger: SEO Authority Score 18 vs. 29–43 cohort. Digital PR, expert bylines, partner content, backlink qualification.
Priority 3 · Future Optional Program · quarterly / 6–12 monthsTrigger: real Maximus demand, thin public proof. Claim registry, flagship case study, buyer FAQs, demo path.
Priority 1 · Future Optional Program · 90-day sprint + advisoryTrigger: D08 defines the map; activation is separate work. Target account dossier, outreach, CRM governance.
Priority 4 · Extra Scope · 90-day activation sprintTrigger: the messaging system (D05) only compounds if the whole commercial team speaks it. Interactive training for the company and the commercial team — shared buyer language, objection practice, playbook adoption.
Priority 5 · Future Optional Program · workshop seriesTrigger: Chapter 54 expansion signal, limited ZA/NG search activity. Country scorecard and localized research before any scaled program.
Priority 6 · Future Optional Program · adjacentTrigger: D10–D12 hand off a roadmap, not an operating system. Keeps the roadmap alive with a monthly refresh and decision log.
Extra Scope — Commercial Agreement Required · ongoing operating program Keeps everything above aliveKanbina intelligence dataset, follow-on module register, 10 Aug 2026. Nothing already promised in the signed SOW is withheld — these operationalize it after Day 90.
Flagged by Kanbina Lumina as a known upcoming touchpoint, alongside the confirmed bauma.de referral already visible in GA4. Event window confirmed: 28–30 Sep 2026, Las Vegas (participant map in the Umbrella Brand section). EKU's attendance and role: pending — not populated with invented specifics.
Detail pendingDashed-underlined terms throughout the document show these definitions on hover.
Whether people who don't know you can find you when they look for what you do. You can be excellent and invisible at the same time.
Making your pages easy for Google to understand and trust, so they show up when someone searches. Not tricks — clarity plus reputation.
The exact words someone types into a search box. If nobody types your words, nobody arrives — no matter how good the page is.
Visitors who found you through unpaid search results — earned, not bought. The opposite of ads.
Branded: they searched your name (they already knew you). Non-branded: they searched the problem. Growth lives in non-branded.
How many times your page was shown on a results screen. Being shown is not being chosen — that's the click.
An industry 0–100 score of how much the web "vouches" for your site through links from other sites. It measures link reputation — not product quality.
Another website linking to yours — a public vote of confidence. More trustworthy voters, higher authority.
How often AI assistants (ChatGPT, Gemini, Google's AI answers) mention your brand when people ask them questions. The new front door.
The AI-written answer box at the top of Google. Being cited there means Google's AI trusts your page enough to quote it.
The instrument panel for your website: who visited, from where, what they did. "Key events" are the moments that matter — like a form submitted.
Google's own report of how your site performs in its search results — real impressions, real clicks, real positions. Ground truth, not estimates.
Your company card in Google Maps and search — name, category, description. Its category teaches Google what kind of company you are.
A position in the market with real demand and no visible owner yet. Finding one is rare; naming it first is the advantage.
Where your page sits in the results list for a given keyword. Position 1–10 is page one; beyond that, almost nobody looks.
This analysis defines the next work — and its validation gates.
The demand-opportunity cross-check does not close in Phase 1: it is the foundation for what comes in Phase 2 and 3. Here is what comes next — and where Kanbina can support, build, or lead.
Six independent evidence streams converge: EKU's technical capability exceeds its digital discoverability. The gap is measurable, named, and closeable.
Every dollar of authority-building and content work has a specific, evidenced target instead of a generic SEO program.
This framework supports Phase 2 messaging, content prioritization, and the Day-90 roadmap sequencing already outlined above.
Working Hypothesis on positioning language and category framing; Confirmed on all cited traffic, authority, and AI-association figures. Full source list available on request.