Best practice for creating an address field for a property survey

[OllyTsimosh]

Hi, I'm relatively new to using ArcGIS Survey123 and I was hoping to get some advise on this.

Here's my situation:
I'm assisting a co-worker with collecting data on vacant commercial properties in the local area that could be then put into a dashboard etc. The surveys are conducted bi-monthly and often the same properties appear (with maybe a change in price, real estate agent details etc).

The issue is that many of those commercial properties are shops/suites or entire floors withing a single building and the Location options in Survey123 are not accurate enough to differentiate between say a shop 1 and a shop 2 within the same building. On top of it, given that this is a rural area, some of the addresses in the geocoder are simply missing or are inaccurate.

My initial (somewhat crude) workaround was to include 2 additional text fields, one for street address and one for sub address (ie Unit 1, Shop 2 etc) that would then be pulled together by a hidden field to form their own unique address. This solution however is highly prone to things like spelling mistakes or inconsistent labelling (ie calling something a Suite one month and a shop another time) when the data is entered by individuals conducting the survey.

If anyone has previous experience building similar types of surveys or have an idea of a good way of recording this type of information, your suggestions would be highly appreciated.

[abureaux]

There is nothing wrong with starting with a geocode field to maybe fill in a few blanks. But you can skip that step if you really want.

Essentially, Id just add separate fields for all the address components. (I would assume that your multi-tenant commercial buildings have unit/suite numbers...)

Here is an example that I use for our radon team. Now, our radon team works anywhere from down-town in large cities to super remote northern territories. aka, the geocoder doesn't always work. But as I mentioned, it's a good starting point:

The user starts by searching for the property in the geocode field. Hopefully they find it and it fills in all the other fields via pulldata(). If that's the case, job done! But, they can also tweak all of the separate address fields as-needed. In the end, I pull all the address data from those separate fields and not the geocode field: