In The News


Bid Metrics in Snow Removal Request for Proposals

Incorporating statistics and weather
By: Sean Hartnett, Manager of Client Services, CSP, Planalytics, Inc.

Kevin Smith, vice president of operations at Ferrandino & Son, has published a very effective and accurate Best Practices document outlining the challenges retailers face when creating an effective/ efficient Request for Proposal (RFP) for snow removal services, a weather-driven service. He specifically outlines how many within the retail industry spend an overwhelming amount of time, effort, energy, and resources focusing on four key items.

1. Scope of work (SOW). Establishing delivery methodology/expectations, governance and reporting.

2. Asset base. Gathering asset inventory data (sq. ft.), site maps, etc.

3. Risk/Liability. Requiring proof of insurance, setting minimums, and defining indemnification clauses.

4. Exhibit pricing. Seasonal, per-event/per-inch, time and materials, etc.

He then focuses on how to effectively and efficiently tie all of these critical pieces of data together to ensure that all of vendors/ contractors are bidding and being measured by the same criteria, which will ultimately lead to a more proficient RFP process and selection.

Again, overall I agree with Kevin’s statements for Best Practices in shaping an RFP for snow removal services.

It is my expert opinion that for a retailer to truly be best in class you must address and understand the most basic/obvious element of snow removal.

The weather was, is, and will always be the fundamental building block of both RISK and OPPORTUNITIES to control costs/budgets, as it drives the need/costs of snow removal services. Retailers that are poised to take advantage of understanding the weather, statistically, will have the ability to create opportunities to capitalize on mitigating risks with their vendors where snow and ice removal is needed, a key benefit in an environment where every dollar counts.

Best Practice
The volatility of the past several winters, featuring mild and extreme temperatures and conditions alike, has brought to light the need for a much more strategic and statistically based approach to incorporating weather information as a fundamental building block in establishing a weather baseline for budgeting and RFP contract selection purposes.

In order to apply a more strategic approach to snow removal budgeting and RFP development, an important first step is to understand how to match contracts (seasonal versus per-event/variable) to the risk profile for each store and for the company overall. Again, weather is THE fundamental building block for constructing an accurate snow removal budget. All too often, retailers ignore the very thing they are attempting to solve for: the impact of actual weather. To ignore statistical weather data can expose organizations to unnecessary risk and can cause them to unwittingly forfeit opportunities to reduce costs, expand margins, and maximize ROI.

In this era of “big data,” to accurately de-risk the budget and contract selection criteria (seasonal versus per-event/variable) during your next RFP cycle, industry leaders are beginning to leverage a much more sound statistical approach based on multiple years of fact-based weather data. This enables retailers to mitigate risk by optimizing and providing visibility into contracting for a specific market-e.g. seasonal, per event/variable.

It should be no surprise that a season’s total snowfall is not the primary driver of the cost of snow removal (though it is obviously correlated). Cost is ultimately determined by a function of both the frequency of “events” per season and how much snow falls during those events (1-3 inches, 3-6 inches, etc.).

How did you begin? What steps did you take? In what order?

Today retailers mine data for weather information as the building block to assemble a weather baseline for budgeting and to support strategic RFP contract selections (seasonal versus per-event/variable rates). Most retailers adhere to a rule of thumb that should a location have an average snowfall total (e.g., 15 inches), the RFP automatically defaults the stores to a seasonal contract versus a per-event contract because this will reduce their risk of spending too much. This may appear to be sound logic on the surface, but in practice, when a sound statistical approach is applied, it often reveals that this “15-inches-of-snowfall- on-average” rule doesn’t provide the best results.

Example scenario: A retail store located in New York City, which averages around 25 inches of snow per year, is seeking snow removal pricing. Resisting the “rule of thumb” mentioned above, the retailer wants to evaluate pricing for both seasonal and per event contracts.

The challenge: How do they measure seasonal versus per-event bids and which contract represents the greatest ROI? Which contract type represents the best value based on the retailers’ risk profile?

Inputs for analysis: Contractors’ pricing (seasonal and per event) and “big” weather data.

A statistically driven analysis will reveal the actual risk to the retailer of a per event costs exceeding the costs of a seasonal bid in any given year. With this information in hand, the retailer is able to make a much more informed and strategic decision based on probabilities, expected ROI and its own risk profile.

In this example, the retailer has received seasonal pricing of $19,500. A per event contract is priced as follows: $525 for 0-1 inch, $900 for 1-3 inches, $1,250 for 3-6 inches, $1,500 for 6-9 inches, $2,100 for 9-12 inches and $125 per inch over 12 inches.

Analysis results: Looking at this specific scenario and the two cost structures, Planalytics calculated an expected annual cost of $11,486 for the per event approach for any given year (based on weather data that considers the number of events and probabilities associated with the snow amount thresholds for this location). Furthermore, the statistically probability of the per event costs exceeding the seasonal quote in this example is only two percent. The analysis shows that “per event” is the more cost-effective alternative; $11,486 expected annual cost versus $19,500 annual cost for the seasonal. Does this approach really work? Looking at a five-year period (2008-12), the seasonal costs in this example were $97,500. The expected per event costs based on weather data and statistical probabilities was $57,430. The actual costs based on the snow events that occurred over the five-year period were $50,625. It turns out that the statistically driven model was a pretty good predictor and the choice to go per event instead of seasonal would have produced savings of $46,875 over the five years.

In the end, these organizations have very little visibility into the risk associated with the volatility of weather as it relates to their decisions, which often results in rogue expenses that exceed budgets. The ability to gather and analyze a large body of relevant weather data is the missing piece of the puzzle when it comes to snow removal contract selection.

By using fact-based information and robust historical data, best-in-class retailers remove the uncertainty of what the weather should be and mitigate the hysteria of last season or the utilization of an arbitrary average. This places those organizations in the best position to select the contract type that is best for them and to more accurately predict how much they should budget to spend. Planalytics ultimately recommends that companies utilize our comprehensive risk modeling approach for preseason planning to improve both budget accuracy and a stronger bottom line performance year-over-year. This proactive approach is the key to maintaining a risk-averse and profitable operation throughout any winter season.

Sean Hartnett is Manager of Client Services, CSP, at Planalytics, Inc., which is available on the web at

share this article