Chief Human Resources Officer / Canada
Decisions that moved work, money and careers
- Colleagues year-round
- 9,600
- Grangemont Farm & Home Inc.; about 11,200 at the spring peak
- Annual compensation and benefits expense
- C$352m
- Fiscal 2025; not the HR operating budget
- Store-manager appointments filled internally
- 77%
- 2025, up from 31% in 2021
Decisions & consequences
01Installing self-checkout without removing the work it moved
In 2023 our finance team proposed self-checkout for the 46 stores with annual sales above C$18 million. The business case removed 196,000 cashier hours a year, worth about C$4.3 million, and assumed the stores no longer needed any of those hours. I doubted it. A large share of our baskets include a twenty-kilogram bag of feed, a roll of fencing or a bale of shavings that nobody scans at a kiosk, and the customer still needs it loaded. Nine of the 46 stores were unionised, and their agreements required notice and discussion of technological change.
I proposed that we install the lanes but reinvest about 60% of the removed hours, for eighteen months, in two roles: yard loading, and advice at the feed and farm counters. We would then measure where the work had gone and decide the rest with evidence. Our chief financial officer, Owen Tessaro, wanted the full saving in the budget, arguing that half-measures would leave us with lanes and the old cost base. Patrice Gendreau, who leads our stores, worried less about cost than about theft at unattended lanes. Martine Ouellard agreed, provided we reported the net saving to the board quarterly.
We gave up most of the first-year saving. Installation ran from March 2024 to April 2025. No cashier was laid off: 117 moved into yard and advice roles at C$1.40 an hour more, and 39 who wanted fewer hours took them. In the nine unionised stores we agreed the role changes with the union before the lanes arrived.
In fiscal 2025 the net saving was C$1.8 million, from 82,000 hours. Self-checkout handled 27% of transactions, almost all baskets under C$60. Complaints about loading waits fell 44%, and front-end turnover in those stores fell from 71% to 52%. Shrink, however, rose from 1.31% to 1.52% of sales, which cost about C$2 million and more than wiped out the net saving in the first year.
Workforce sections are now part of every store technology case. What I would do differently is involve loss prevention from the start. When they redesigned the lanes in 2026, with an attendant post facing the exit and weight checks, shrink fell back to 1.39%. I should have asked for their design at the start.
Bargaining with the cost model on the table
The agreements covering 410 employees at our Drummondville distribution centre and 140 at the feed mill expired on 31 December 2022, after a year in which consumer prices had risen by close to seven per cent. The union asked for 19% over three years. Our initial mandate was 10% over four. The distribution centre supplies all 56 Quebec stores and our 11 New Brunswick stores, and because Quebec's labour code restricts the use of replacement workers during a strike, a stoppage in April or May would have emptied those stores of feed and seed in the weeks that carry nearly a third of their annual sales of both.
I proposed showing the union's bargaining committee our cost model for the distribution centre before we made a first offer: cost per case shipped, a wage survey of six regional warehouses, the volume plan and the cost to the company of a six-week stoppage in spring. Our external labour counsel warned that disclosure could never be withdrawn; our general counsel feared leaks to competitors. Martine Ouellard agreed after I proposed limits. The committee would see distribution-centre costs and company-wide margins, but not store-level results, and its members would sign a confidentiality undertaking.
We gave up the advantage of knowing more than the other side, and some room to plead that we could not afford more. The committee correctly challenged two of the six warehouses in our survey, and we replaced them.
Both agreements were ratified in April 2023, by 81% at the distribution centre and 77% at the mill. They run for five years, with increases of 6.5%, 4% and then 3% a year, and a scheduling clause that posts weekly schedules fourteen days ahead and caps mandatory overtime outside the peak at six hours a week. Labour cost per case rose 8.7% in 2023; after we re-slotted the warehouse it was 3.9% above 2022 by 2025. Turnover at the distribution centre fell from 34% to 18%, and there was no stoppage.
I would change two things. The mill followed the distribution-centre pattern, although its labour market is different and its real shortage is millwrights, not production staff; it deserved its own conversation. And because only the committee saw the numbers, members heard rumours of secret figures. Information sessions for all members would have cost little and prevented that.
Growing store managers in the towns that needed them
In 2021 we had 29 store-manager vacancies across 168 stores. The average vacancy took 143 days to fill, only 31% of appointments were internal and 41% of the external hires left within two years. Relocation packages averaged C$34,000, and the managers we relocated left more often than those we did not. Most of our stores are in towns of fewer than twenty thousand people, where a manager is also a known figure in the community, and a vacancy of five months meant an assistant manager running the store without the authority or pay to do it.
I proposed a hub-store model. Thirty-four larger stores would each carry a second assistant manager in development, chosen from stores in the surrounding area and paid to cover holidays, leave and vacancies in three or four neighbouring stores. Every acting week would be recorded, and a candidate would normally have ten weeks of acting cover before appointment. We would stop paying relocation for store-manager appointments unless the candidate asked to move.
The model cost about C$3.6 million a year. Owen Tessaro questioned paying for thirty-four positions that did not run stores. Several regional managers objected that hubs would take their strongest people, and some store managers resented losing an assistant to a neighbouring store. Patrice Gendreau supported the model on condition that regional managers chose the candidates. We also accepted slower appointments at first: in 2022 the average time to fill rose to 151 days, because we waited for internal candidates.
By 2025 the picture had changed. Of 26 store-manager appointments, 20 were internal. The average time to fill was 49 days, two-year retention of managers appointed in 2022 and 2023 was 86% and relocation spending fell from C$612,000 to C$96,000. Annick Thébeau, appointed at Victoriaville in 2024, had covered four stores for a total of fourteen weeks before taking her own.
The board's committee now sees the bench twice a year by region and language. I would change two things. We launched in Ontario first and reached Quebec a year later because the development material was in English. And we first drew candidates mainly from front-end supervisors, when the strongest turned out to be feed and farm counter leads who already knew the customers.
Making French the working language of our Quebec stores
In mid-2022, after Quebec amended its Charter of the French language, I asked my team to review how our Quebec operations actually worked in French. The findings were uncomfortable. Of 310 training modules, 42% existed only in English. The scheduling application due for company-wide launch in spring 2023 had no French interface for managers. Nine of fourteen Quebec district and regional management postings listed English as a requirement, and our monthly operations calls ran in English. Our Quebec stores were compliant on signage and much less so in the work.
I proposed that Quebec be treated as a French-first operation rather than a translation queue. We would re-author training in French with Quebec store colleagues, not run it through a translation service. Every English requirement in a Quebec posting would need a written reason based on the duties, and company-wide meetings would have simultaneous interpretation. We would also delay the scheduling application by seven months until its French version had been tested in stores.
The chief operating officer disagreed. The scheduling application was expected to save C$2.2 million a year in labour, and a delay meant losing more than C$1.2 million of that. Some Ontario regional leaders thought interpretation would slow meetings. The one-off cost was C$2.4 million, plus about C$380,000 a year for interpretation and translation.
Our board approved the plan in October 2022. We re-authored 130 modules with forty Quebec store colleagues and retired the rest. Twelve of fourteen management postings dropped the English requirement; the two that kept it work daily with English-speaking suppliers. Training completion in Quebec stores rose from 58% to 93%. When the scheduling application launched in Quebec in late 2023, 96% of store managers were using it within eight weeks, against 81% in Ontario. Internal appointments to store manager in Quebec rose from 3 of 11 in 2021 to 12 of 14 in 2024.
French-language review is now a step in every HR system decision. What I would do differently is start with Quebec store managers rather than a compliance checklist; for six months we treated the work as a legal exercise. I also forgot New Brunswick. Four of our eleven stores there serve mainly French-speaking communities, and their staff received English-only material until 2025.