A ROM is usually a single point estimate or a range but still related to a single instance. As such, precision doesn't apply. Accuracy can and if a ranged ROM is provided, so long as the final actual falls within the range, you could say it was accurate. With a single point ROM, unless it happened to land very close to the final actual, it wouldn't be accurate.
Estimates only need to be as accurate as what is required to meet your needs. Developing detailed estimates cost more time and money so use your resources wisely.
In the early phases of projects, ROM estimates generally fall into 3 categories:
1) Slam dunk - Even using simple conservative numbers, the business case or schedule estimate looks very good and is worth pursuing further. This might be followed by a more accurate estimate to enable committing resources to the project.
2) No chance - There is no chance of returning a positive business case or a schedule that meets our needs. We don't need to spend any more money showing how bad of an idea it is. The ROM estimate is enough to stop wasting resources.
3) Sharpen your pencils - Using a ROM, the concept may be feasible, but more detailed estimates are required due to the uncertainties in the original estimate. What you learned in the ROM will often tell you where you need more accuracy as well so you can focus on the important variables, and not the ones that don't significantly affect the outcome. Saving Changes...
Senior Projects Manager | Field & Marten AssociatesNew Westminster, British Columbia, Canada
ROM estimate are usually +/-50% accurate but could also be +/-%75 accurate depending if the organization has done similar past projects, or not. They are very high level. Saving Changes...
Peter RapinSubject Matter Expect; Project Delivery| Independent ConsultantOntario, Canada
Maybe I am reading this wrong but ROM for me is Random Order of Magnitude. The word Random does not make me think of precision nor accuracy. I don't like the terminology at all.
My thinking is that a preliminary estimate should be based on past experience and lessons learned (same thing). That is why a client typically retains a cost engineer or such to generate this costing. Using a professional should increase the precision (repeatability based on data base and process). The greater the contingency the higher the precision. A professional should be able to achieve a reasonably high level of precision at +/- 50% as suggested.
...
1 reply by Kiron Bondale
Dec 18, 2022 5:47 PM
Kiron Bondale
...
Peter -
I've always seen ROM's full form as "Rough order of magnitude". Not random, but certainly not something you'd want to commit to...
Maybe I am reading this wrong but ROM for me is Random Order of Magnitude. The word Random does not make me think of precision nor accuracy. I don't like the terminology at all.
My thinking is that a preliminary estimate should be based on past experience and lessons learned (same thing). That is why a client typically retains a cost engineer or such to generate this costing. Using a professional should increase the precision (repeatability based on data base and process). The greater the contingency the higher the precision. A professional should be able to achieve a reasonably high level of precision at +/- 50% as suggested.
Peter -
I've always seen ROM's full form as "Rough order of magnitude". Not random, but certainly not something you'd want to commit to...
Kiron
...
1 reply by Rami Kaibni
Dec 18, 2022 5:52 PM
Rami Kaibni
...
Kiron
I agree - Ive always seen and used it as Rough Order of Magnitude and we only base minimal decisions based on it.
A ROM is usually a single point estimate or a range but still related to a single instance. As such, precision doesn't apply. Accuracy can and if a ranged ROM is provided, so long as the final actual falls within the range, you could say it was accurate. With a single point ROM, unless it happened to land very close to the final actual, it wouldn't be accurate.
Kiron
Right said Kiron. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
My recommendation is taken a closer look to Barry Bohem work, mainly "Cone of Uncertainty". Regards Saving Changes...
Mark WarnerProject Manager| AURATucson, Az, United States
ROM stands for “rough order of magnitude,” and, as its name implies, it’s rough, meaning not very accurate. When you’re first starting a project, this is often the best you have available to start initial planning. This is fine—provided you (a) fully communicate the inaccuracy of the number to your key stakeholders; and (b) ensure there’s enough initial contingency assumed/allocated to cover the large error bars on the estimate. As things move forward, and your estimates improve, this contingency number can be reduced accordingly.
The thing people forget is that every estimate, from the most inaccurate (e.g., ROMs and SWAGs) to the significantly more accurate (e.g., detailed bottom-up estimates and vendor quotes), are still little more than “guesses.” Until the work is complete and the bills paid (i.e., risk reduced to zero), they will remain guesses, and they will require some kind of contingency allocation to cover the baked-in uncertainty. Saving Changes...
Stéphane ParentSelf Employed / Semi-retired| Leader MakerPrince Edward Island, Canada
I typically use rough order of magnitude (ROM) along with level of effort (LOE) estimates.
The purpose of the ROM is to give you an idea of the size of the project for decision purposes: should we proceed or not? There is an understanding that the final actual project size will be within plus or minus 50% of the ROM estimate.
The purpose of the LOE estimate is for work planning. You've decided to go ahead with the project and now you are trying to size the work packages and the schedule. There is an expectation that more information is available at this stage to help with the estimates. It is expected that the final actual size of the work will be within plus or minus 20% of the LOE estimate. Saving Changes...
Community Champion
