What changed in 65 days, and what the next 25 decide
The market already trusts EKU's engineering when it knows the name, and almost never meets EKU when it describes the problem. Phase 1 named that gap and got the language approved. Phase 2 built the system that closes it. Phase 3 needs five inputs from EKU to prove it with numbers.
Day 65 of 90Session · Thu Sep 3, 2026Working document
503 / 744AI answers describe EKU correctly when the prompt names EKU. The reputation exists. Jul 2026, defined prompt set.
0 / 5,176AI answers to buyer-problem prompts, no brand named, that mention EKU. The gap this program closes.
99.2%of clicks to ekupd.com come from people typing the EKU name. Search Console, 14 Jul to 13 Aug 2026.
The five-minute read · for Edward
New since Aug 25, the results, and what I need from you
New since the team session
Field Terms delivered per territory (Sep 1) · the master workbook consolidated: one source of truth, 169 dated layers · this readout, bilingual, plus the Q4 2026 + 2027 operating plan (top-right button) · competitive sweep extended: 21 brands in 5 tiers, master battlecard live, MTU move into the space flagged · a new buyer-language evidence layer (what buyers actually type) · every deliverable now passes an automated quality gate before you see it.
Results on the board
US organic keywords 14 → 29 in twelve months (Jul 2025 to Jul 2026) · referring domains 208 → 248 (16 Jul to 12 Aug) · clean, like-for-like search baseline established: 297 clicks/mo, every future claim measurable · AI systems describe EKU correctly 503 of 744 times by name, 0 of 5,176 by problem: the gap is now a measured starting line, not an opinion · EKU is the only cohort member whose visibility is rising.
What I need from you today
1. 20 minutes on the claims list (D05): sign or correct, unblocks the website and all sales material. 2. Maximus case numbers by Sep 15. 3. The eight pipeline averages (no CRM access needed), or name who answers them. 4. Green light for the GA4 events before Sep 11. 5. Pick the October door (A, B, C or D) before the Germany travel.
Story room · the whole engagement in five scenes
Open the scenes in order. Each one links to its full chapter.
1
A company the market trusts, and cannot find
Where we started on July 1
Ask about EKU by name and the market answers correctly: 503 of 744 AI answers describe EKU accurately. Describe the problem EKU solves, without the name, and EKU appears 0 times in 5,176 answers. Engineering was never the problem; language and visibility were.
EKU's territory is Power Asset Readiness, proven by Chip-to-Cloud, spoken in the buyer's problem language. "Mixed-brand generator fleets" replaced the retired open-ended compatibility claim, ratified unanimously by the team on Aug 25.
Messages, channels, content engine, partner map: D05 to D08, plus the measurement baseline that makes every future claim checkable. The sign-off list on the table today moves it from working-final to final.
The big three engine brands defend their names. The controls cohort defends its keywords (DEIF 439 vs EKU 19). The data-center layer is being bought. Nobody occupies the mixed-fleet orchestration center EKU owns. A timing question, not a capability question.
What was built is a living system, and living systems need an operator. Four honest doors: EKU alone, EKU executes with guidance, operated with Kanbina, or an in-house hire. Deciding the structure this week, before the Germany travel, keeps October seamless.
Known by name. Invisible by problem. Measurable from here on.
0 of 5,176AI-assistant answers to buyer problems (no brand named) that mention EKU. When the prompt names EKU: 503 of 744 answers describe it correctly. Jul 2026, defined prompt set.
99.2%of clicks to ekupd.com come from people typing the EKU name. Google Search Console, clean export, 14 Jul to 13 Aug 2026.
18 vs 29 to 37Authority Score, EKU vs the specialist cohort (DEIF, ComAp, Deep Sea, Basler). An SEO / backlink metric only, 12 Aug 2026.
What you are seeing
Think of the market as a library. EKU's file is in it and it is accurate: ask the librarian about EKU by name and the right answer comes back 503 times out of 744. But buyers describe a problem ("controls for generators from different brands"), and in 5,176 of those questions the librarian never opened EKU's file. None of this is about the product; it is about language and visibility, which is what this program repairs.
The 90-day journey
Three phases, one system
Done · validated Jul 31Days 1 to 30 · FoundationD01 to D04 delivered and validated live with Edward. Language decided. Google identity repaired.
Working-final · Aug 19Days 31 to 60 · InfrastructureD05 to D08: messages, channels, content engine, partner map. Six added-value artifacts delivered.
Gated on 5 EKU inputsDays 61 to 90 · Growth architectureD09 to D12: revenue map, 12-month roadmap, priority stack, handover. Day 90 = Sep 28.
What you are seeing
Phase 1 answered "what is true and what do we call it". Phase 2 answered "what do we say, where, and to whom". Phase 3 answers "what is it worth, in what order, and who runs it after Day 90".
Phase 1 · Days 1 to 30 · Delivered and validated Jul 31
The diagnosis, and the three decisions it produced
D01 · Commercial Landscape
Delivered
Who competes where: OEM ecosystems, controls specialists, complete-power players, packagers. The empty center: mixed-brand plus system-level.
D02 · Buyer Segment Map
Delivered · profiles at Working Hypothesis
Roles, pains and research language by segment. Upgrades to validated as the field questionnaire comes back.
D03 · Digital Authority
Delivered · clean baseline 13 Aug
Search, site, authority and AI-answer baselines, each dated and scoped. Every future claim measures against these.
D04 · Validated Positioning
Delivered · approved Jul 31
Power Asset Readiness as the umbrella narrative; "mixed-brand generator fleets" as the buyer descriptor; Chip-to-Cloud as the proof.
What you are seeing
Four documents, but the value is the three decisions taken in the room on Jul 31: Power Asset Readiness is the story; "OEM-agnostic" is retired from customer language because nobody searches it, replaced by "mixed-brand generator fleets"; and Maximus is the growth focus while idle management stays protected as revenue. On Aug 25 the whole commercial team confirmed the language unanimously.
Already fixed during Phase 1 (added value, at no added cost)
Google Business Profile recategorized: "Electronics company" removed; power-generation identity restored. New description approved. US verification visit still pending.
Search Console export repaired: a filter had contaminated the previous data. A clean, like-for-like baseline now exists (14 Jul to 13 Aug 2026).
Site analytics collecting since about 4 Aug 2026. Named key events specified; configuration is the pending IT step.
Defensive domain list prepared; acquisition authorized to Amanda on Jul 31.
Priority web copy delivered for deployment; "OEM-agnostic" flagged for removal from /maximus.
Phase 2 · Days 31 to 60 · Working-final v1.2 · Aug 19
The operating system: what we say, where, to whom
D05 · Messaging System · messaging house, four role messages, packager sales play, page copy. Gate: 20-minute claims sign-off.
D06 · Channel and Digital Architecture · every channel gets one job (packagers primary); measurement plan with named key events.
One operating system usable Monday morning, every number sourced
D07 · AI and Search Content Framework · three layers: capture demand, name the gap, own the category. Six briefs ready to build.
D08 · Partner and Channel Map · built around the channel truth: about 70% of revenue flows through packagers. Enablement kit spec.
What you are seeing
Phase 2 turned the diagnosis into things people can use on Monday morning: the sentence a seller opens with, the question that sorts a prospect in 30 seconds ("how many generator brands are in your fleet?"), the pages the website needs, and the kit a packager can repeat without us in the room. "Working-final" means content complete and sourced; what remains is EKU's sign-off on a short list of claims.
Added value delivered in Phase 2 (beyond the contracted scope)
Artifact
What it is
Feeds
White Space Matrix
The competitive map with the empty center, placed with measured data
D01 · D04
Readiness Concept Map
Power Asset Readiness as four faces: control, integration, power management, lifecycle
D04
AI / Search Opportunity Map
Where EKU can realistically win AI-era visibility, and the countable KPI for it
D03 · D07
Learning Center Module Map
Six education modules, from "What is Power Asset Readiness" to a plain-language glossary
D04 to D08
Event Language and Player Orbit Map
Yotta 2026 (Sep 28 to 30, Las Vegas) mapped as language intelligence
D01 · D05 · D07
Backlink and Citation Plan
Seven sequenced moves to grow authority where it counts
D03 · D07 · D08
Evidence Pack
Every figure with source, date, scope and limitation, in one workbook
D03 · D12
AI Baseline Tracker
37 prompts locked as the wave-1 baseline; first capture wave scheduled
D03 · D07
Global team session kit (Aug 25)
Story Room, Battlecard Arena, field questionnaire, Field Terms per territory (Sep 1)
D02 · D05
GEO On-Page Pack
Eight Q&A blocks and six definitions, ready to paste, with structured data
D07
Evidence board
Every number here has a source, a date and a scope
Read each block on its own. Different instruments are never averaged or compared with each other.
1 · AI assistants: the file exists, nobody opens it
503 / 744Answers that connect EKU to the right capability when the prompt names EKU
89Distinct technical descriptors AI uses for EKU (controls, cross-manufacturer integration, remote monitoring, switchgear)
0 / 5,176Answers to buyer-problem prompts (no brand) that mention EKU
11AI Overview keyword slots where EKU already appears (10 Aug). Small, real, the toehold.
Source: Kanbina Lumina AI-answer testing, defined prompt set, Jul 2026; AI Overview count 10 Aug 2026. Tool- and set-specific; not market share.
2 · Authority Score: how much weight the web gives each site
EKUControls specialistsComplete-power and adjacentEcosystem-scale benchmark, not like-for-likeMTU: named by EKU leadership, measured in the next pull
Source: Kanbina Lumina market intelligence dataset, normalized pulls 10 to 17 Aug 2026. Authority Score is an SEO / backlink metric only. *Caterpillar is a broad equipment domain; a scale reference, never a one-to-one comparison.
What you are seeing
Authority Score is like a credit score for a website: how many reputable sites point to it. EKU sits at 18; the controls specialists sit between 29 and 37. It moves slowly and only with genuine third-party references. It says nothing about product quality or brand strength in the field.
3 · AI Visibility: share of voice inside AI-generated answers
EKUControls specialistsComplete-power and adjacentEnterprise platforms, scale signal onlyMTU: named by EKU leadership, measured in the next pull
Source: Kanbina Lumina market intelligence dataset, comparison scope, 14 to 17 Aug 2026. EKU's 14 is the comparison-scope value; a separate dashboard surface reads 0 under a different configuration. The two are never merged. AI Visibility and Authority Score are different metrics and are never size-compared.
What you are seeing
If Authority Score is the credit score, AI Visibility is how often a company gets mentioned when AI assistants answer questions in this space. EKU and the specialists sit close together; the giants at the bottom are a reminder of scale, not targets. The program's job is to move EKU within its own group, measured the same way each time.
4 · The search baseline, and what has already moved
297 clicksAbout 7,000 impressions · 4.2% click-through · average position 9.1. Search Console, 14 Jul to 13 Aug 2026.
27 → 29US organic keywords, 15 Jul to 12 Aug 2026. Organic drift before the content engine starts.
208 → 248Referring domains, 16 Jul to 12 Aug 2026. Backlinks 902 → 923. More inventory is not yet more authority.
568 → about 1Page-one impressions per month on "idle management" and the clicks they produce. Visibility exists; the words do not convert it.
US organic keyword count by month, Jul 2025 to Jul 2026 (domain-rank history, Kanbina Lumina market intelligence dataset, Aug 2026). 14 in Jul 2025 to 29 in Jul 2026: real, positive, and still small enough that one good quarter moves it.
What you are seeing
297 clicks a month is the honest starting line. The line is the number of US search terms EKU ranks for: roughly doubled in a year, from a low base. The "568 to about 1" tile is the sharpest example in the dataset: people already see EKU on page one for "idle management" and do not click, because the page speaks company language, not theirs. Fixable with words, not budget.
5 · Buyer language: measured demand, and where EKU ranks today
Buyer search term
Cluster
US searches / mo
Difficulty
EKU rank today
smart PDU
Energy management and distribution
480
7 (low)
not ranking
generator monitoring system
Data intelligence and remote ops
320
2 (low)
not ranking
generator control panel
Controls and automation
260
2 (low)
not ranking
microgrid controller
Energy management and distribution
260
28 (moderate)
not ranking
power plant control system
Controls and automation
170
3 (low)
not ranking
smart power distribution unit
Energy management and distribution
170
8 (low)
#28
remote generator monitoring
Data intelligence and remote ops
140
0 (open)
not ranking
black start generator
Blackstart and auxiliary energy
70
14 (low-moderate)
not ranking
Maximus controller (branded)
Product
40
1 (open)
#18
idle management system
Fluid management and efficiency
20
0 (open)
not ranking
Source: Kanbina Lumina keyword-demand analysis, 10 to 13 Aug 2026. Volumes are modeled monthly US searches per exact term: directional estimates, never revenue forecasts. Where the specialist cohort ranks and EKU does not (31 Jul 2026, US desktop): "ats controllers" 140/mo, "generator paralleling controller" 70/mo, "diesel engine controller" 50/mo. Domain-level: Deep Sea 337, ComAp 395, DEIF 439 keywords vs EKU 19.
What you are seeing
Each row is a phrase real buyers type, how many per month, and how hard page one is. "Difficulty 2" means the door is almost open; "28" means established players are there. EKU appears in two rows only. The strategy follows the table directly: open doors first (controls, monitoring), then the moderate one that proves the category (microgrid controller), and teach Power Asset Readiness on top.
The competitive field · one map, three layers
Who the buyer meets on the way to EKU
Three layers, never mixed: brand gravity, the measured controls cohort, and the data-center power layer. Each carries its own instrument and date.
1
Brand gravity · the engine "big three"
Caterpillar: measured, ranks on 9 of the shared top 10 buyer terms (31 Jul 2026). MTU (Rolls-Royce) and Cummins: named by EKU leadership as the other dominant engine platforms; scale references, scheduled for the next benchmark pull.
2
The measured controls cohort
DEIF 439 · ComAp 395 · Deep Sea 337 organic keywords ranked (US) vs EKU 19, and EKU is the only cohort member whose visibility is rising. Gap analysis 31 Jul 2026.
3
The data-center power layer · being bought now
Schneider, Vertiv, Generac and Caterpillar occupy every measured data-center power term. EKU is absent from all of them, and one direct competitor is already paying for ads on Google and on the new ChatGPT inventory.
Buyer term
US searches/mo
Who occupies it today
data center power solutions
1,300
Schneider #22 · Vertiv #44 · Generac #36
data center backup
390
Generac #13 · Caterpillar #18
generator paralleling switchgear
110
Schneider #3 · Caterpillar #11
The big three own brand-side searches. EKU does not compete with their engines; EKU makes mixed fleets of their engines operable. That is why the problem-side terms, not brand terms, are EKU's territory. The language has a price, and someone else is paying it.
What buyers actually type (the newest evidence layer)
Controller searches carry a brand
"deep sea / comap / deif genset controller". Buyers search brand plus product, not the bare category: comparison and intercept pages beat generic category content.
Monitoring searches carry an OEM
"kohler / cummins / generac / cat generator remote monitoring". The buyer searches through their engine brand, then hits the mixed-fleet wall: each phrase is an intercept-page opportunity.
Paralleling is a question goldmine
"can you parallel generators of different sizes / brands", "do parallel generators need to be the same". The different-brands questions ARE the mixed-fleet problem, in the buyer's own words.
Source: Google autocomplete capture, Sep 2026, category seed terms; Kanbina Lumina market intelligence dataset. Directional language evidence, not volume figures.
What you are seeing
The window: the big three defend their brands, the controls cohort defends its keywords, the data-center layer is being bought, and nobody occupies the mixed-fleet orchestration center EKU actually owns. This is a timing question, not a capability question.
What is still open · Days 65 to 90
Phase 3 is built from five inputs only EKU can provide
D09 · Revenue Opportunity Map
Working draft Sep 11 · final Sep 18
With the CRM export or the eight averages: shaped by the real pipeline. Without them: demand-side only, clearly labeled.
D10 · 12-Month Roadmap
Sep 16 to 22
Sequence, owners, KPIs and dependencies. Needs the continuity decision: who operates the system weekly from October.
D11 · Commercial Priority Stack
Sep 18 to 22
P0 / P1 / P2 with an explicit "not now" list. Needs leadership tradeoffs, not more data.
D12 · Presentation and Handover
Sep 23 to 28 · Day 90
Board-ready package, source register, evidence annex, owners per asset. Delivered as an upload-ready package.
Six decisions for today
1 · Claims sign-off (D05): claims list, integrated-controllers list, publishable Chip-to-Cloud diagram. 20 minutes, sign or correct each claim. Unblocks the website release and all sales material.
Edward · this week
2 · The proof asset: the committed Maximus case with the customer mentioned, with numbers by Sep 15. The power generation gap is where a case earns most.
Edward / Amanda · by Sep 15
3 · Revenue map input (D09): CRM export, deal sizes, win/loss, or the eight averages from memory. With it, D09 lands with the real pipeline shape.
Edward + Finance · by Sep 11
4 · Google Business Profile US verification: one in-person visit at the Spring office. Natalia can complete it if no owner is available this week.
Spring office / Natalia
5 · Measurement events (GA4): the named events are specified in D06; configuration is an EKU IT task. Before Sep 11, Day 90 reports business signals; later, visits only.
EKU IT (Amanda) · by Sep 11
6 · Continuity before the Germany event: who operates the system weekly from October. Agree the structure for the next six months today.
Edward · today
What you are seeing
None of these asks for more analysis. They are the five pieces of information the analysis cannot generate on its own, plus one decision about ownership. Each has an owner and a date so the list closes itself.
Also open, with owners (not decisions for today)
Field questionnaire: 0 of 5 replies received. Battlecards enter production "field input pending" with the first replies (sales team via Amanda).
Partner list and relationship history; confirmation of whether packager relationships may be visible publicly (Edward, D08).
LinkedIn analytics access; role assignment deferred until data exists (Amanda, D06).
Data center as a served segment: confirmation before the data-center brief and buyer profile go live (Edward).
Owners, capacity and budget band per initiative for the roadmap (Edward + Amanda, D10 / D11, Day 75 review, week of Sep 14).
Site release checkpoint with Amanda, end of September.
The next six months · who operates the system
What was built is a living system. Living systems need an operator.
Everything delivered stays with EKU and is documented and transferable, whatever is decided here. The open question is who runs the weekly engine from October.
Macro · quarterly rhythm
Narrative and roadmap governance · competitive benchmark pulls (adding MTU and Cummins) · category leadership: creating events, tech talks and trade-magazine features · claims governance so nothing unsourced ships · quarterly team training refresh.
Micro · weekly and monthly rhythm
AI-answer panel re-run and register · search and site monthly read · content briefs into production with QA · battlecard updates as field input arrives · site release and Confluence upkeep · event follow-through (data-center event in Las Vegas, mining tech talk, Germany).
Four ways to run it (all four are workable)
A
EKU runs it alone
Internal ownership
Needs a named owner (roughly half to one full role), tool licenses, and a realistic 3 to 6 month ramp to rebuild the 90-day context. Viable if the owner exists today.
B
EKU executes, Kanbina guides
Advisory cadence
EKU hands do the work. Kanbina keeps the measurement, governance, and monthly command review, and coaches the team. The system stays honest; speed depends on EKU capacity.
C
Kanbina operates with EKU
Full operating cadence
Everything in B, plus production: briefs, battlecards, event assets, biweekly refresh. Zero ramp, full continuity of context from day one of October.
D
In-house hire
Employment route
A full-time commercial growth role (40 flexible hours, benefits, admin access to the commercial stack). Consolidates ownership in-house; carries a hiring cycle plus onboarding time.
Commercial reference points (open when the conversation gets there)
Option
Structure
Monthly
Market reference
B · Guided
6 months, 3+3 checkpoint, monthly command review, one benchmark pull per month, claims governance, team coaching
$9,200
Growth-marketing engagements run $7,000 to $15,000/mo; senior consulting hours $110 to $200
C · Operated
Everything in B plus content production, battlecards, event assets, biweekly refresh
$12,370
Below the fractional-CMO band ($10,000 to $20,000/mo) with more execution included
Houston marketing-director benchmark: median $194,092/yr, typical range $178,800 to $208,200 (Sep 2026). Total employer cost with benefits typically 1.25 to 1.4 times salary.
For scale: option C over 12 months is about $148,000, roughly 55 to 60 percent of the typical in-house total cost, with no ramp and the 90-day context intact. Sources: MarketerHire consultant rate report 2026; Salary.com Houston benchmark, Sep 1, 2026. Public references, not quotes.
What you are seeing
A decision with four honest doors, not a pitch. The only door that costs momentum is leaving the question open: the site release, the Las Vegas event and Day 90 all land in the same four weeks, and the system needs an operator before then. Deciding the structure this week, before the Germany travel, keeps October seamless.
An impression is EKU appearing in someone's search results. A click is that person choosing EKU. Many impressions and few clicks means people see the name and the words do not persuade them.
Authority Score
A credit score for a website, built from how many reputable sites link to it. It moves slowly. It is an SEO metric; it says nothing about product quality.
AI Visibility
How often a company is mentioned when AI assistants answer questions in this space, within a defined comparison. Different tools and settings give different numbers, so the setting is always named.
Modeled search volume
An estimate of how many people in the US type a phrase each month. Directional, useful for choosing which phrases to own. Never a sales forecast.
Like-for-like
A comparison only counts when both numbers come from the same instrument, geography and settings, and are both dated. Every "before and after" on this page passes that test; anything that does not is left out.
Claim labels
Confirmed validated with EKU or measured directly. Working Hypothesis evidence supports it, EKU validation pending. Needs Validation stays out of external material until confirmed.