American here, so I can't comment too much on how robo-advisors work in Canada, but in the US they suck. The reason is that the fees, although small, are large enough that they destroy any additional performance the robo-advisor might give you.
In fact, here it is straight from the horse's mouth:
https://www.wealthfront.com/historical-performanceWealthfront straight up doesn't beat the benchmark. But their main pitch is they automatically maximize tax savings. But in the U.S. anyway, the maximum amount you can deduct is less than the fees they charge, at least for medium-large portfolios. And in fact, MMM is doing a real-time experiment that shows Betterment is lagging the benchmark:
http://www.mrmoneymustache.com/betterment-vs-vanguard/I have to eyeball the graph a little bit, but it looks like the Robo-investor underperformed by about $25,000 over the past four years (I'd guess the index is at about $200K and the roboinvester at $175K, give or take). That's real money.
To be fair, the Robo-advisor claims it has harvested $35,000 in capital losses, which would be roughly $10,000 in tax savings. Since $10,000 is less than $25,000, you can see right away that it ain't worth it.
But it is worse than that! In the U.S. you can only deduct $3,000 in capital losses per year. Over five years, that's a maximum of $15K in deductions, which translates to maybe $5K in tax savings. Giving up $25K to get $5K is a bad deal.
But it is even worse than that! By harvesting the losses now, you have lowered the basis which increases your tax in the future. Because multiplication is associative, it doesn't matter mathematically if you pay the tax now, or later. The end result is the same (all things being equal of course).
Their secondary pitch is that they automatically rebalance so you don't have to. But studies have shown that monthly re balancing is no better than annual. Some studies say five year periods is fine. And in fact, rebalancing based on time periods might not be the best way to think about it anyway:
https://www.kitces.com/blog/best-opportunistic-rebalancing-frequency-time-horizons-vs-tolerance-band-thresholds/In short, you don't need or want a robo-investor (sorry for the U.S-centric nature of this post).