That thin layer of workble soil is pretty much SOP for developers and many landscaping companies........it is the least expensive route, no one sees it and it is just enough to get things going. But really insufficient to keep a lawn or any other plantings healthy long term.

I consider 6-8 inches of friable, enriched and well draining soil about ideal. You can get by on less but the better the soil conditions, the better the lawn establishment (deeper root penetration and therefore better drought tolerance) and general plant growth.

Although it is a not very sexy aspect of landscaping and barely even visible, good soil prep is not anything I would cut corners on. It can make or break a garden. And since you appear to be working with a blank slate, now - before anything else happens - is the best time to deal with it.

Gardengal is thinking of the ideal situation. If you can afford it, then have the top 6-8 inches hauled away and replaced. Just to scope this, 6 inches equals 18 cubic yards of material per 1,000 square feet. That is about the capacity of one 18-wheeler. How many square feet are we talking with this project?

Never EVER rototill in preparation for grass. If you can get a tractor to do it, then it sometimes works. The problem is that tillers grab soft soil and sand and dig way down making a hole filled with fluffy soil. They also hit roots and rocks and buck up out of the ground leaving an underground hill with much less fluffy soil. When you level it all, you end up with different amounts of fluffy soil everywhere. When it all settles in, the holes will be holes and the hills will be hills. You'll have a bumpy surface to mow.

Never EVER mix compost into the soil. Compost contains plenty of undecomposed wood. When wood starts to decompose, the fungus that decomposes it will steal nitrogen from any source available. When you fertilize, ALL the nitrogen goes to wood decomposition. You can use compost, just don't bury it under the surface.

Getting the grading right is most important. Don't bring in any topsoil or any other soil unless you have low spots to fill. If you bring in enough soil that it sits up higher than the concrete walkways or curbs, then you will have soil washing out onto the concrete for the next 20 years. The soil level must be below the concrete level. The soil should drain out toward the street or drainage ditch and never toward any buildings. Your builder should have done this for you.

What if you only have 2 inches of "good soil?" It can be fixed. First of all ensure you have a good microbial variety by applying a scant amount of compost. You already have the variety, but most people want better assurance. Compost has it all. Scant means 1 cubic yard to cover 4,000 square feet. Toss it out willy nilly. That's good enough. Then I would spray the yard with shampoo at a rate of 3 ounces per 1,000 square feet to help water penetrate deeply into your relatively poor soil. Any clear shampoo works. I use generic baby shampoo. Some folks like the apple scented shampoo from the dollar store. Water this in deeply - at least 1/2 inch and preferably 1 inch as
measured by tuna or cat food cans placed around the yard. This might
take 8 hours depending on your sprinklers. Deep means real deep.

Then scatter something like corn meal, alfalfa meal or pellets (rabbit chow), soybean meal, corn gluten meal, cottonseed meal, or whatever you can find at your feed store for under $20 per bag. The application rate is 15 to 20 pounds per 1,000 square feet. This is where the magic happens. This is the feed for those microbes you injected with the compost. These grains can be moistened but do not need to be watered in like a chemical fertilizer.

This regimen of compost, shampoo, and organic fertilizer will improve your soil to a very deep level. The next level would be to get a good soil test from Logan Labs and post the results here on this forum for help interpreting it. Logan Labs has the best quality control over the long term, and their most basic ($25) test is all you need. You can get similar tests done from your local university, but to get the same level of detail, the add-in testing will run you upwards of $60 at most and $100 at some. Logan Labs can keep their price low because every test gets all the testing, so they buy chemicals at a much higher volume discount than other labs.

The next thing to do is get roots in the ground. If you live in the south and are planning to use a hybrid bermuda, you should wait until June to start. Bermuda really likes HOT soil. Most other grasses can be sodded in the spring with pretty good results. Seeding grasses in the spring anywhere but the Pacific Northwest seems to invite crabgrass to the lawn. Fall is a better time to use seed for grasses like fescue, rye, and Kentucky bluegrass. If that is your situation, you can try a cheap seed this spring and nurture the weeds and sparse grass until then learning how to water, mow, and fertilize for several months. Then nuke it and start over. People do this all the time.

Hmmmm.....feel free to take the above with a grain or two of salt :-) You do not need to "scrape away" 6-8 inches of soil - that is just the depth of friable, enriched soil you need to work with. Adding to what is there is perfectly fine. And you can certainly till and add compost. Not all compost has a high percentage of yet not fully decomposed bark...and it really doesn't hurt if it does. Just bump up the N!! And since the soil will get leveled and then rolled both before and after seed application, there will not be any high, fluffy spots or dips and valleys.

As a landscape professional I oversee the installation of a couple dozen new lawns each year and this is a tried and true method for my area.....not sure why it should be any dfferent elsewhere except for the type of grass used.

(the lawn care forum guys tend to be very dogmatic in their approach - lawns are really dead easy to install and maintain)

Definitely some interesting ideas. I'm not worried about roots and tilling. There's one large tree in the center of the back yard and its root structure is well below current grade. It seems like unless I add 6" of good soil to the top of the subsoil/basement spoilage currently out there I'll need to amend down somehow.

I ordered a soil sample kit from the Michigan State University extension, which is probably the best in Michigan. I'm in zone 6a and all we do is kentucky blue grass (blends). I think the trick will be finding an installer that will be willing to do more than the bare minimum.

Adding to what is there is perfectly fine. And you can certainly till
and add compost. Not all compost has a high percentage of yet not fully
decomposed bark...and it really doesn't hurt if it does. Just bump up
the N!! And since the soil will get leveled and then rolled both before
and after seed application, there will not be any high, fluffy spots or
dips and valleys.

Hilarious. Sure you CAN do all this stuff. I was trying to avoid complications starting on Day 1 after the sod guys leave and continuing for years to come. Gardengal and I have been getting along for so long that I forgot what she and I disagreed on way back when. The high and low spots are underground and caused by the rototiller bucking up over the hard spots and digging in to the soft spots. When you level that off at the surface, you are guaranteed to have deeper and shallower fluffy soil everywhere. Plus when you till you destroy the soil structure that takes years to develop. It is best to level the surface by scraping the high points off and filling the low spots. Fill, roll, fill, roll until the filled area seems the same density as the surrounding soil. It's not the same as structure, but it's better than 3 to 8 inches of varying subsurface profile.

Just bump up the N!! ...for the next 10 years while the wood decomposes. Seriously. That's years and years of disappointment looking at the yellow lawn even though you bumped up the N.

If you improve the soil from the surface, the plants' root systems and healthy microbes will fix the soil down deeper. You fix the soil at the surface with the chemicals recommended by the soil test, by repeated applications of organic fertilizer over the years, and deep infrequent watering.

I ordered a soil sample kit from the Michigan State University extension, which is probably the best in Michigan. ...you would think Ohio State would be the best in Ohio, but they are not. Logan Labs is way at the top of the list across the US, with UMASS coming in a close second, and all the rest falling far behind fast. A bunch of lawn guys from another forum sent the samples around and came up with this. There is a private lab in Texas which is much better than Texas A&M, but still not at the top of the list. But if you don't mind doing it, you can still post the results here for comment. There won't be any labism from the Logan Labs purists. Promise.