I guess it boils down to looking at all the factors and making a call. Time and Money are two factors, but there are others.
A big factor to consider is risk/liability. What are the consequences of the data being off? For example, is the data being used in a Life Critical Application or other purpose where there are high expectations that the data be solid and consequences of inaccuracy when the data is relied upon and fails severe? An evaluation will need to be made of this. Not just the liability in the legal sense, but also the risk that poor data will cause other side effects, like for example poor location data used in a construction project causing delays and rework as things turn out not to be quite where the data seemed to indicate they were.
Along these lines, another aspect to consider is standards. Are there accuracy standards that are required that are not being met? External standards, Agency standards, Internal standards, Client standards? Will the data in its current "not perfect" state meet the standards? If it doesn't meet an expected standard, that needs to be mentioned. Organizations I've been with have occasionally had issues with subcontractors where they failed to meet the data standards we specified, which had major financial consequences for them as laid out in our contract agreement (and consequences for us as the client as this often led to project delays and rework).
Also, there is the less tangible risk - if there is an expectation by clients (internal and external) that the data is good and if its not quite up to snuff from their perspective, will your group/organizations reputation take a hit? Will this impact the long-term ability to get clients/work with other organizations?
As part this evaluation, also keep in mind the realities of any project, that there almost never is enough resources to do it all perfectly. That's not to say one shouldn't strive to do better, but it may be what is currently available is sufficient to do the tasks at hand. Keep in mind there can be a dimension present in that it was known that the data was not great but that it would be workable, so a project was bid out with that knowledge, thus the reluctance of the organization to spend more money to improve the data. This can be a major factor in private sector work.
So make a frank evaluation. Maybe the data is good enough as is? If not, "spend time and money now" to improve the data or "spend time and money later" when things go awry?
Whatever decision is made, another item to consider is metadata. Is there an explanation of the data limitations in the metadata? Do you have a disclaimer? It might be worth documenting the known limitations to help prevent other users from running into trouble as they may not see the data issues and consequently have issues.
Anyways, those are some things to consider.
Chris Donohue, GISP
