Lab IA JB · September 2026
Ganon Group in the age of AI
What we have understood about your group, what AI answer engines say about it today, and what AI can change — with a timeline for every project.
Prepared for Ronen Ganon · Jeremy Bouznah — Lab IA JB. Everything that follows comes from your websites, the business press and tests run in Hebrew on three engines on 2 and 3 September 2026. Nothing is estimated.
What we have understood about you
Four divisions, a dozen brands, a group that is growing fast.
Zol Stock
stores, the country's second-largest discount chain. 50% owned by the group. HOM'S and Home Style for the home.
Shai Market
neighbourhood convenience stores launched that year, alongside the HaKatzavim MiYerushalayim butcher chain and Netto Hisachon in Tsfat and Beit Shean.
GG Electric
across three logistics centres, a new site in Kiryat Shmona, and the group's commercial refrigeration brand.
Nofei Bereshit
homes in Tsfat, 13 buildings. Limassol Heights in Cyprus. GG Towers on the former Nescafé plant: 164 homes, announced in June.
Ganon Group Holdings — קבוצת גנון אחזקות — led by Ronen Ganon, around 550 employees according to your website, plus a community relations division alongside the three businesses. Sources: ganon-group.com, Globes, TheMarker, Ynet, 2 and 3 September 2026.
The moment you are in
Your scale is changing. Your data foundation is not.
stores after the planned combination with Multi Retail
billion shekels of annual turnover cited for the combined group
of Multi Retail Group's capital stays on the stock exchange: public reporting from closing day
Agreement of 12 August 2026 for 72.3% of Multi Retail Group (ACE, Autodepot, Bitili, Urban), listed in Tel Aviv since 2021, for ₪173.5M, subject to the Competition Authority and Ace Hardware International. Your three websites are measured in three different ways: no data can be aggregated. Sources: Globes, Calcalist, TheMarker.
What has changed in search
Your customers no longer search. They ask.
A list of links
The customer types "cheap appliances", scans the results, compares, chooses. Being on page one was enough to exist.
A single answer
They ask ChatGPT or Perplexity and get three to five retailers. If you are not among them, you do not exist for that buyer.
On 2 September 2026 we asked, in Hebrew, the questions your customers would ask Perplexity, ChatGPT and the first page of Google results. The next three slides show what they answer.
Zol Stock is cited first. A position to defend, not to conquer.
A watch that continuously asks the engines what they say about Zol Stock and triggers a response as soon as a reservation appears — the engine already voices some about certain warranties. No tool on the market does this job.
Perplexity, Hebrew question "cheapest discount chains for the home in Israel", 2 September 2026, 1:23 pm. Two of the seven chains cited were offline the next day: a real position, and a fragile one.
Zol Stock versus Max Stock, Stock Home, Stock Online
The awareness exists. Nothing converts it.
An online store born ready for AI agents — the chain has no legacy system to reconcile, while its competitors will have to adapt a shop designed for humans. An app and a loyalty scheme that finally say who buys what.
Volumes on a square-root scale. Stock Online's open checkout is provided by Shopify: within reach of any recent store. Findings of 2 and 3 September 2026.
The only shopping cart in its market. Invisible to machines.
Product pages the engines understand, then a search that understands a question asked in everyday language. Nobody in this market has done it, and the cart already exists: it is the group's best effort-to-result project.
None of the five competitors has structured product data, customer reviews or a file for AI engines. On the "appliances" query of 2 September, GG Electric is absent from all three engines. Benchmark of 3 September 2026, volumes on a square-root scale.
Real estate division
Twenty developers cited. No Ganon project.
developers recommended by Perplexity for new residential projects: Africa Israel, Azorim, Aura, Shikun & Binui, Ashtrom, Tidhar, Gindi, Rothstein…
Nofei Bereshit — not cited
Limassol Heights — not cited
GG Towers — no web page
Aura shows 153 project pages and 58 city pages; Tidhar follows its buyers on a portal. Your group website: five short project pages, no city page.
City pages, a presence on the portals the engines cite (Madlan, Yad2), and a buyer portal that follows every purchaser from the first payment call to handover of the keys, without relying on a salesperson's memory.
Perplexity, Hebrew question, 2 September 2026, 1:15 pm. Sitemaps of the five developers read on 3 September. This is the division with the widest gap, and therefore the one where the first progress will be the most visible.
A detail that shows from the outside
Three of your group's addresses do not respond.
limasol.ussl.co.il
The Limassol Heights link points to no server. A buyer, a partner, a journalist: error page.
www.zolstock.co.il
Certificate error: the browser shows a security warning before the page even loads.
nofei-bereshit.co.il
"Under construction" page since 2024, with a plan of 185 homes waiting behind it.
Verifiable in thirty seconds, tested on 2 and 3 September 2026. It is not the most serious point in this presentation; it is the simplest to fix, and the first thing anyone looking for you sees.
Your digital set-up today
Three websites, three worlds. No data can be aggregated.
- Host A, behind Cloudflare
- A single advertising pixel
- Management: S-Tech
- Host B
- A measurement tool obsolete since 2024
- Provider: Studio Puenta, 2019
- Host C
- Modern measurement, but isolated
- No provider credited
This is not a quality problem: each site does its job at its own scale. It is a scale problem: at the moment of integrating several chains, nobody has the job of making these systems talk to each other. Your current providers do hosting and websites; we come in where nobody has stepped in yet.
Logistics and food
Two divisions where there is nothing to undo.
Three centres, no shop window
25,000 m² and a new site in Kiryat Shmona. What the trade does today: real-time flows, replenishment forecasting, supplier portal. Your position: to be assessed in the field.
No web address
What the trade does: localised store pages, traceability, waste and freshness tracking, neighbourhood loyalty. A blank page to write the day you decide to.
Build at the right level from the start
Structured for the engines, connected to group management, with no technical debt to pick up later.
Assisted replenishment, phone-guided stock counts, traceability and waste tracked in the butcher shops, supplier portal with deadline alerts. Every building block is qualified with your teams before it is proposed.
The heart of the system
A single view of the group, before the first listed-company report.
Revenue and margin by entity, consolidated cash position, indicators ready for your shareholders.
Stock turnover by store, projects unit by unit, logistics flows.
Performance compared by chain, a single customer view whichever chain they shop at.
Consolidated headcount by entity and division, multi-site scheduling.
Consolidated volume by supplier, group negotiation — "a table, a grater or a lamp", as your partner puts it.
Regulatory obligations, product traceability, reputation monitoring from AI engines.
Four building blocks make it possible: a unified product master, accounting consolidation, a unified customer base, harmonised measurement. Built during integration, they cost one project. Rebuilt once the shareholders are at the table, they cost far more. We connect what exists; we only develop where necessary, and we tell you where.
What AI changes in your businesses
Four building blocks that touch your teams and your customers.
A customer recognised at one chain is recognised everywhere in the group. Reconciling customer bases entered differently is one of the rare places where AI brings what nothing else does.
For Zol Stock first: catalogue, nearest store, promotions, loyalty card. Finally knowing who buys what, and how often.
Scheduling, headcount by entity, tracked recruitment and training, connected payroll: the foundation for absorbing the arrival of new teams.
What ChatGPT, Perplexity and Google say about each chain, measured over time, with a response triggered as soon as a reservation appears.
We do not yet know your internal systems. Every building block starts with questions to your teams — how a stock-out is reported, how a schedule is built, how a buyer is followed — then a proposal: integrate a market solution when it does the job well, develop when a real interest justifies it.
Where you can arrive first
Three areas of lead, not catch-up.
A Zol Stock online store born AI-ready
No legacy system to reconcile. Its competitors will have to adapt a shop designed for humans. The delay becomes a head start.
Structured group data from the integration onwards
Linking the holding to each of its chains during the acquisition, rather than rebuilding afterwards.
Continuous monitoring of what AI says about you
No player observed, in retail or real estate, does it systematically today.
The first horizon — online store, customer account, loyalty, content — is a quick catch-up with proven building blocks. The second — files for AI engines, structured data, reviews — is where the immediate gap is decided. The third — purchases made by the customer's own agent — is the only one that lets you get ahead.
How long it takes
From early October 2026 to autumn 2027.
Scoping, access, first fixes: the three addresses, the files for the engines.
Fixes delivered, harmonised measurement across the sites. Projects per chain under way.
Projects per chain largely delivered. First tier of the cockpit, AI watch in place.
Second tier: operational and financial view. Advanced technical integration.
Third tier: HR, purchasing, compliance. Consolidated, auditable group data.
The anchor point: from the closing of the Multi Retail deal, expected by the end of 2026, the group consolidates a listed company. The first tier arrives with its first quarterly accounts. Several timelines depend on things we do not yet know — the systems of the newly combined chains, access to the right people. We would rather say so.
One last thing
What we found is what everyone finds.
Nothing in this presentation comes from your teams: everything comes from your websites, the press and the AI engines, exactly as a customer, a partner or a future shareholder looking for you sees it.
If this portrait does not match who you are — a business poorly described, a brand missing, a figure that does not look like you — the gap itself is the first project. It is what people land on.
We experienced it while preparing this document: GG Electric presents itself as a commercial refrigeration brand, yet the engines file it nowhere. Telling machines who you are is the first building block of everything above.
To go further
The full dossier remains at your disposal.
The site-by-site audit, the tests run on the three engines, the panorama of innovations in each of your businesses across three horizons, and the operational detail of each division.
Method and sources. Your official websites, the Israeli companies registry, the business press (Globes, Calcalist, TheMarker, Bizportal, Walla, Ynet), tests in Hebrew on Perplexity, ChatGPT and Google's first page, direct examination of your competitors' websites by division. Engine answers change from one session to the next: all the more reason to measure them over time. Findings of 2 September 2026, re-checked on the 3rd.
Jeremy Bouznah — Lab IA JB · Clarity. Strategy. Impact.