I agree, yet still I use one. It's the convenience factor. For the small fees they pay, they implement the ETF strategy I would otherwise use. When I'm adding money, there's zero work or thinking.

Now, if I had $250k or more I might change my mind- there's a certain point where yes, it's worth my time to do this myself. But I'm not there yet.

If you setup a spreadsheet with your asset allocation, and blanks for your relevant accounts (30-60 minutes), it should take about 5-10 minutes to plug in current numbers when it's time to add money, or every once in a while if you want to rebalance. If you're just starting accumulation, rebalancing with contributions might make the most sense.

It doesn't make sense to me to pay AUM % for sometime that takes an hour to setup, and minutes a quarter. If they wanted to charge $10 per deposit, maybe. (I also don't see much value in the automated tax loss harvesting, especially if they're not doing a great job avoiding wash sales, wtf?)

I tried that with robinhood, but that few minutes every month to dollar-cost average in + dealing with whole shares felt like a big distraction.

I switched to m1finance where you can set a "pie" e.g. 80% SPY, 20% bonds and then just add money. They do once-daily investing, and automatically add new money or dividends to even out the pie.

Now I once a month I just dump "extra" money in without needing to do any calculations.

Do you have a spreadsheet template? Finding the efficient frontier and the relevant numbers seems hard?

It sounds like you're having trouble with asset allocation, more than figuring out how to implement your allocation. Nevertheless, here's a description of my spreadsheet.

I'm using a variant of the three fund portfolio. So I've crafted a target percentage for stocks and bonds, and the target percentages for each of the funds I'm using for that. Then you add things like your traditional 401k balance (which I've put in bonds), your roth 401k balance (in stocks for me), and maybe you've got a couple of those, it's easier to have one line for each, so you can run through the online accounts and put it in.

Anyway, so on the left side of the sheet, I've got the list of all those funds, the target %, the current balance, and then the target balance. On the right side, I have my new contribution -- most of my contributions are coming from equity based compensation, so I have a bunch of stuff over there to help me set aside the right amount for taxes (I could have a side rant on that).

Target balance is computed for taxable stocks and bonds overall by doing (total balance + contribution) times allocation% - amount of that in tax advantaged.

Then, for each fund in taxable, I take the overall balance for stocks or bonds and use the ratio of allocations within that class to compute the target for that fund (if I have 30% for bonds, and 21% is for fund A, and 9% for fund B; whatever I got for taxable bonds times 21 divided by 30 is the target for A.

Things to note: a) depending on the size of your taxable and tax advantaged balances and your overall asset allocation, tax efficient fund placement may dictate what goes where. For me, my tax advantaged has clear choices, so each type (tax deferred, roth) gets fully allocated to one type of fund. Also, you want to avoid using "substantially identical" funds in taxable and non-taxable, to avoid potential wash sale issues that are hard to track and have unfortunate consequences (if you do a wash sale in taxable, it's fine, you don't get the loss booked, but your cost basis is preserved; if you sell for a loss in taxable and buy substantially identical in tax advantaged, the loss is disallowed but the basis is not transferred, so you just lose that forever).

If this isn't very helpful, I can probably make a clean copy of my spreadsheet as a template; I would share mine directly, but it's hard to know what hidden data is in there.

If you just select 80/20 there's no mean variance optimization going on...

If you want to do mvo, there's tons of free tools on the web. Or is very easy to build one yourself in Excel. You need the solver and at least a year of historical data for each name.IEX cloud has a fantastic free API for this.

Especially with tax loss harvesting and rebalancing, I'm suspicious of how roboadvisors do when you factor in the bid-ask spread. At least with ETFs, you can be confident someone like Blackrock isn't getting ripped off by limit orders or front-running trades.