The headline number is store count, the real story is data
Michael Hill reported on September 1 that its Canadian business posted CAD 174.2 million in revenue, up 7.3 percent, with same-store sales up 7.0 percent and the first eight weeks of fiscal 2027 already running at 9.8 percent same-store growth. CEO Jonathan Waecker framed the market opportunity plainly: "In Canada, we are just facing into a market with an incredible amount of opportunity." The company plans to grow its Canadian footprint from 81 stores to a target of 85 to 90, with five high-volume locations across Toronto, Vancouver, Calgary, and Edmonton scheduled for modernization in fiscal 2027.
That is a modest expansion for a retailer posting these growth numbers, which is exactly the point. A specialty jeweler that wanted to chase revenue the easy way would open dozens of new doors and lean on real estate to carry the growth story. Michael Hill is instead growing same-store sales through personalization and digital investment first, and treating new stores as a secondary lever rather than the primary engine. For a category with high average order values and low purchase frequency, where a single customer might transact once or twice a year, that is the more capital-efficient growth path, and the same-store and online numbers back it up quarter after quarter.
Where the growth is actually coming from
Canadian e-commerce grew 22 percent in fiscal 2026, more than double the 10 percent growth rate of Michael Hill's global online business. That gap is worth sitting with: whatever Michael Hill is doing in Canada specifically to drive digital growth is outperforming its own average by a wide margin, and it is happening in a market where the company is simultaneously investing in personalization infrastructure. Gross margin improved 20 basis points to 60.3 percent even as the company expanded its Made For You personalized product line to more than 15 percent of sales, evidence that the margin gain is coming from personalization itself, functioning as a margin-accretive lever rather than a discount-driven growth tactic dressed up as innovation.
The bespoke jewelry service, which lets customers design custom pieces, has expanded to more than 40 stores, a meaningful footprint for a service model that requires trained staff and a workflow most jewelry chains never bother building out. Combined with clienteling tools that give frontline staff a data-informed view of a shopper's history, past purchases, and stated preferences, Michael Hill is building the kind of high-touch, high-margin service model that independent jewelers have historically owned as their core differentiator, and doing it with a data infrastructure layer a single-location independent retailer simply cannot match at any price point.
The AI inventory project is the less glamorous but more telling investment
Personalization gets the marketing attention, but Michael Hill also disclosed an AI-enabled inventory planning project underway, alongside plans to expand buy-online-pickup-in-store capability. Inventory planning is unglamorous compared to a shopper-facing AI assistant, but for a jeweler carrying high-value, low-velocity SKUs across dozens of locations, better demand forecasting directly reduces the two costs that erode jewelry retail margins fastest: markdowns on slow-moving stock and stockouts on fast-moving pieces during peak gifting seasons.
Pairing that inventory work with expanded BOPIS is a coherent, sequenced bet rather than two unrelated line items on a technology roadmap. Accurate store-level inventory visibility is a prerequisite for reliable pickup promises, and a retailer that gets this wrong trains customers to distrust its online stock indicator, which quietly kills conversion on exactly the high-consideration purchases jewelry represents, where a customer who drives to a store expecting an item in stock and finds otherwise rarely gives the retailer a second chance that day. Michael Hill sequencing these together suggests the inventory project is being built with omnichannel fulfillment treated as a design requirement from the start, rather than a feature bolted on after the forecasting model was already in production.
Why this is a better proof point than most personalization case studies
Most retail personalization case studies report engagement metrics, click-through rates, session length, that are easy to move and hard to tie to revenue. Michael Hill's numbers are harder to argue with: comparable EBIT grew 16.3 percent to CAD 21.9 million, faster than the 7.3 percent revenue growth, meaning profitability improved faster than sales. When a personalization and data investment shows up in margin expansion rather than just top-line growth, that is a stronger signal the investment is working as intended rather than simply buying growth through discounting or paid acquisition.
It also matters that this is a mid-size specialty retailer, not a hyperscale platform with unlimited data science headcount. Jewelry is a considered, infrequent purchase category where personalization has to work with sparse per-customer signal, one or two purchases a year rather than weekly grocery runs. A retailer proving personalization ROI in that harder data environment is a more credible proof point for other specialty and durable-goods retailers than another case study from a company with daily transaction volume to train on.
What this means for specialty retail technology leaders
If you run technology for a specialty or durable-goods retailer with infrequent purchase cycles, jewelry, furniture, appliances, outdoor gear, Michael Hill's sequencing is worth copying: build the inventory and personalization data layer before committing capital to store expansion, and measure success in margin, not just traffic. The temptation in specialty retail is always to open more doors when comps are strong. This result argues that the data layer compounds in a way square footage does not, and it de-risks the eventual store expansion by ensuring new locations inherit working personalization and inventory systems rather than starting from zero.
The AI-enabled inventory planning project is the piece worth watching longest term. Personalization gains are visible quickly in e-commerce conversion, but inventory accuracy compounds slowly and shows up as margin over several quarters, which is exactly what Michael Hill's EBIT growth outpacing revenue growth suggests is happening. Retailers evaluating their own AI roadmap for next year should treat inventory and demand forecasting as at least as high a priority as any customer-facing shopping assistant, even though it will never generate the same press coverage.
