A Case Study Over Lunch

Harvard Business School is famous for its case method. Harvest, across the river in Harvard Square, has hosted dinners for the school’s conferences and programs for years. After a morning exploring the neighborhood, it seemed an appropriate setting for a second, smaller case study: what would our lunch cost when paid with discounted inKind credit?

We had been visiting Harvard’s athletic grounds, looking at the field hockey and rugby facilities and the main football stadium. I was also interested in locating the business school, opposite the athletic complex. From there, we wandered back toward Harvard Square and the train, checking the inKind map for somewhere to eat.

The choices along our route included ramen, sandwiches and a gastropub that would not open until early evening. Harvest was the main sit-down lunch option available to us. That made it a useful follow-up to ⁠my New York inKind case study, which ended with the obvious condition that the stated value of the prepaid balance depended on finding suitable restaurants for the remaining balance.

Harvest has been serving Harvard Square since 1975. Its connection with the business school extends beyond geography: HBS schedules place conference and program dinners there as far back as 2016, with repeated appearances through 2025. For one HBS conference on international prices and income, the World Bank sponsored the welcome dinner at Harvest. A restaurant where economists discuss price comparisons was a fitting place to examine one of our own.

The inKind listing categorizes Harvest as “Business Dining,” and the crowd at lunch on Tuesday, Oct 20, appeared to have taken that description seriously. Everyone seemed dressed up. I did not see evidence of a formal dress code, but the atmosphere suggested that making an effort was encouraged.

I never attended HBS, but I spent four years at a college in a university town and scarcely remember seeing anyone in a suit. This was a different university-town lunch crowd. You do not see that concentration of dressed-up diners everywhere even in central London on an ordinary Tuesday, either.

The food was pretty good, and the portions were enormous. Our shared shrimp cocktail contained four exceptionally large shrimp arranged over ice. My fried chicken sandwich arrived cut into two substantial halves alongside a container overflowing with fries, while the hummus bowl was a suitable main in its own right.

We ordered the shrimp cocktail to share, a fried chicken sandwich, a hummus bowl, two pilsners and two glasses of prosecco. Food and drinks came to $119, with $8.33 in tax bringing the restaurant bill to $127.33.

I added a $23.80 tip, exactly 20% of the subtotal before tax. The app deducted the restaurant bill from my existing inKind Cash and charged the gratuity separately to my Amex Gold.

Harvest’s listing advertised 20% back, just as Mercato’s had in New York. It did not apply to this payment. Spending covered by prepaid inKind Cash does not earn that additional return, and gratuities are excluded too. That distinction is explained fully in the earlier case study. Here, the relevant benefit was the discount already embedded in the balance I brought to lunch.

I had bought $1500 of inKind Cash for $999 during an earlier promotion. I deliberately timed that purchase to coincide with Citi Nights on my Citi Strata Elite, and the transaction earned six ThankYou points per dollar.

The $999 charge produced 5994 points. At my working valuation of 1.5 cents each, those points were worth $89.91, an effective 9% return on an already discounted purchase price. That reduced the effective acquisition cost of the $1500 balance to $909.09, or approximately 60.6 cents per dining dollar.

Applying that acquisition cost to the $127.33 used at Harvest gives:

$127.33 × $909.09 ÷ $1500 = $77.17

The $23.80 tip went onto Amex Gold, where a 4× Membership Rewards return is worth 6% at the same 1.5-cent valuation. Allowing approximately $1.43 for those rewards puts the effective cost of the tip at $22.37.

The result was almost identical to New York’s 34.0% reduction. That was unsurprising: we were spending credit acquired through the same promotion and paying a similar proportion of the total as a separate tip.

These figures measure the reduction from the restaurant bill. Paying the whole amount directly with a rewards card would also have earned points, so the incremental advantage over that alternative would be smaller.

Another interesting development was apparent on the app’s purchase screen. The offer displayed when I checked was $999 for $1275 of inKind Cash, rather than the $1500 I had received earlier.

The price paid was unchanged, but it bought $225 less dining credit. The app described the newer offer as a 28% bonus. That is a bonus relative to the amount paid, rather than a 28% discount on the credit’s face value: $999 for $1275 represents a 21.6% discount before card rewards.

For a consistent comparison, assume another $999 purchase earned the same 5994 Citi ThankYou points. That would again produce an effective acquisition cost of $909.09, but this time spread over only $1275 of credit.

Each dining dollar would therefore cost ~ 71.3 cents. The same $127.33 restaurant bill would consume credit that cost $90.79 to acquire, rather than $77.17. The tip calculation would remain unchanged.

The displayed offer could still produce a worthwhile result, but it would make this identical lunch $13.62 more expensive on an effective-cost basis. Nothing about the menu, restaurant payment or meal would need to change.

This newer-purchase column is a scenario, not a second transaction I completed. It assumes another purchase made during Citi Nights would qualify for the same six-point restaurant rate as my original transaction.

At Harvest, the inKind Cash balance did what we wanted it to. It covered a restaurant that was open, on our route and suitable for the lunch we were looking for. We did not have to cross town or rearrange the day to redeem it.

Whether to buy more credit is a separate decision. The earlier promotion remains part of the cost of my existing balance; a less generous offer does not retrospectively change what I paid. It does change the calculation for replenishing it.

Both calculations assume the full purchased balance is eventually used. Under inKind’s terms, the paid portion is redeemed first and does not expire, while the promotional portion expires three years after purchase. Allocating the acquisition cost evenly across meals is how I account for the combined purchase. Leaving promotional credit unused would clearly worsen the eventual result.

That makes the map and the purchase screen equally relevant. One shows whether there are restaurants that fit the trip. The other shows what it would cost to fund them.

Our Harvest visit produced a pretty good lunch, four enormous shrimp and another successful use of the balance. The $151.13 payment had an effective cost of ~ $99.54 under the economics of my earlier purchase.

The second case also exposed the limit of repeating the first result. At the displayed purchase price, even assuming the same card rewards, that lunch would have cost ~ $113.16. Harvest supplied a suitable place to spend the credit; the earlier promotion determined how much value we received from doing so.

New to inKind? My referral link offers new members $25 off a qualifying meal of $50 or more. It cannot be combined with another inKind balance or offer. I receive inKind Cash Back after a qualifying payment; the referral reward varies with current promotions.

A Manhattan inKind Case Study