I'm a Millennial early adopter who finds grocery shopping to be both stressful and irritating, mainly due to the presence of other human beings, who clog up the aisles and checkout lines. Why aren't I buying all of my household's food online yet? Isn't that what Instacart and Amazon are vying for me to do? Yep, and so far they're falling short.

Prompting a large-scale change in consumer behavior, the way Amazon did for books and Uber did for taxis, requires a substantial increase in convenience and a substantial decrease in annoyance, without being too much more expensive. That's what it takes to overcome inertia and the force of long-ingrained habit.

Right now, the state of online grocery shopping is nearly as stressful and irritating as its IRL counterpart. The space may seem saturated, and on-demand food delivery of any kind is certainly a tough space to compete in. But there's still room for startups to actually get it right.

A better browsing experience is one way to do that. Scrolling through the many options in a virtual supermarket provokes paralysis analysis, on top of needing to worry about price parity.

For starters, no website or app has replicated the breezy experience of walking down an aisle, visually scanning the options, and zeroing in on what you want. Online grocery shopping is more along the lines of staring at an incomprehensible wall of slightly varying toothpaste tubes, except extended to every product category.

Aside from its eye-roll-worthy branding, recent entrant Brandless gets the closest to my ideal online grocery experience. Every product is generic and there's only one of it. As Business Insider explained, "There are only two real choices to make on Brandless's website: What do you want, and how much of it do you want?"

Unfortunately, Brandless has $9 shipping until you hit $72, and only offers non-perishable items. Amazon's Prime Pantry is afflicted by the same downsides -- in particular, the $5.99 shipping charge removes the irresistibility of normal Prime. You have to buy a lot before the shipping prices feel worth it, since they're on par with the cost of day-of delivery from Instacart and its ilk.

Any service that involves a third-party shopper visiting a normal grocery store on my behalf will be some combination of too expensive and not reliable enough. I can't trust the shoppers to replace unavailable items with anything approaching my own judgment, and the databases they're working with aren't accurate enough to know whether specific products are still going to be available by the time my order is being packed. (This actually applies to Safeway's own delivery service as well, for some godforsaken reason.)

It's galling to pay for my carefully curated grocery order to be delivered, and then still need to do supplementary shopping because much of what I chose wasn't available. Good Eggs solved this problem by operating its own warehouses and directly managing its own inventory. The company is essentially what you'd expect if Whole Foods had been founded by techies in 2011.

But Good Eggs has struggled, both with a "whole paycheck" reputation to match its apparent inspiration, as well as with reaching financial sustainability, despite having raised roughly $50 million in venture capital. The company only operates within the San Francisco Bay Area, and even then doesn't cover many cities.

Younger startup Farmstead, which started operating in May 2016 and left stealth on Thursday, aims to avoid Good Eggs' financial woes by using artificial intelligence to eliminate waste. The algorithm determines when more stock is needed, which is especially key for the perishables -- milk, eggs, produce -- that are Farmstead's backbone. (Even Amazon, with all its logistics prowess, hasn't figured out how to optimize inventory: "Amazon Fresh has lost money from spoilage at more than double the rate for a typical supermarket," Bloomberg reports.)

Farmstead relies on small distribution hubs and offers a limited, heavily curated selection of products. It picks the best of the best, says CEO Pradeep Elankumaran, occasionally expanding the options because of customer feedback. The flagship "micro-warehouse" in San Mateo, California, maintains 850 SKUs, give or take, and doubles as the startup's HQ. The second location in San Francisco has 600 SKUs.

Taking a note from Good Eggs and Imperfect Produce, Farmstead encourages customers to set up weekly deliveries, and waives delivery fees for subscribers. The startup has made 15,000 deliveries in the past 12 months, and has 70 percent customer retention. It's been growing 30 percent month-over-month, according to Elankumaran, and recently raised $2.8 million from investors including NFL alum Joe Montana.

Is Farmstead the answer to my online grocery shopping woes, my desire for an actual one-stop shop that doesn't require leaving my house? Well, no, because the company don't serve my city yet. (Imperfect Produce does, thankfully, and I'm a happy subscriber.) If Farmstead did, I might never leave my house.

The combination of a non-overwhelming product selection, the confidence that I'd receive the specific products I ordered, and reasonable delivery fees for short time windows sounds just right. Let's hope that Farmstead can keep it up.