MAISY AI-Powered Energy Apps Blog
We periodically publish short notes to address topics relevant to our AI-Powered Web widget applications.
Artificial Intelligence (AI) App Estimates Household/Business
Energy Use and Emissions with Just a Few Dwelling Unit/Household or Business/Operating Inputs,
May 3, 2023
The new age of artificial intelligence has opened an analytic toolbox that offers significant advantages in estimating energy use and emissions of households and businesses compared to traditional tools.
The MAISY Energy Cost & Sustainability Assessment App uses just a few dwelling unit/household or business/operating characteristics in an AI application to estimate energy use and emissions for any dwelling unit or commercial building in the US. The App also provides a comparison of user energy cost and emissions with peer businesses/households in the same location along with savings available with improved energy efficiency targets.
The App applies MAISY Energy Use Residential and Commercial Customer Databases which include energy use and emissions to match the App user’s dwelling unit/household or building/operating characteristics with a “neighborhood” of peer households/businesses. The databases include more than 7 million households and businesses across the US. This AI matching process uses a non-parametric k-nearest neighbor (KNN) machine-learning analysis. Database records are identified as belonging to the user's neighborhood with a weighted distance measure that includes several dozen of the user’s characteristics. The AI process determines individual characteristics weights and applies a statistical refinement process to develop energy use and emissions for the user’s property based on energy use and emissions data from its closest neighbors. Energy customers within the neighborhood define a distribution of energy costs and emissions that provides the user with a comparison of their energy cost and emissions relative to peer neighbors along with estimated cost and emissions savings that can be achieved with energy efficiency improvement targets.
This AI approach significantly improves individual business/household energy use estimates compared to traditional energy use estimating models as illustrated in a comparison of the approaches in the following sections.
Traditional Individual Energy Use Modeling Approaches
Engineering models estimate energy use for prototype buildings. These models apply assumed building shell characteristics, internal building loads (e.g., people, lights, etc.) and hourly weather data to calculate energy use based on engineering relationships for prototype buildings. Model results can be adjusted to reflect information from a sample of actual energy users; however, the prototype nature of the results can make matching actual energy use to real-world energy customers difficult. This modeling methodology is used by the National Renewable Energy Laboratory (NREL) to provide estimates of energy use and hourly loads for prototype buildings.
Statistical models (referred to as conditional demand analysis models – CDA ) use statistically estimated energy use estimates rather than engineering calculations. Conditional demand models apply regression analysis to tease out the impacts of individual appliance holdings and occupant characteristics and then apply the regression parameters to estimate energy use expected to occur with any set of characteristics.
End-use models apply estimates of energy use for individual end uses (heating, water heating, lighting, etc.). The end-use estimates can be developed with a combination of engineering analysis, statistical CDA analysis or metered data and can reflect variations in end-use energy as a function of household/business characteristics like number of household members. Total building energy use is calculated as the sum of the individual end use energy estimates. DOE’s Energy Information Administration applies this methodology to provide end-use energy use estimates from its residential and commercial energy customer surveys.
Accuracy Problems With Traditional Individual Energy Use Modeling Approaches
All traditional models apply “fixed-parametric” methodologies which means that model parameters are constant and, after model development, are applied in the same way to each household or business energy use estimation within a building or geographic segment.
Traditional fixed-parametric models often result in significant errors in estimating individual customer energy use because of “missing variables “and/or “missing values.” Missing variables errors occur because some variables important in determining energy use were not included in the models. Missing values errors refer to errors that occur because values for variables in the models are not available for the individual household or business.
Engineering models are subject to both “missing variables “and “missing values.” For example, these models use assumptions on the number of dwelling unit air changes per hour reflecting leakage through doors, windows, etc. Since this variable is unknown for most applications, the model assumptions are unlikely to be accurate for any individual dwelling unit and can significantly impact heating and air conditioning energy use estimates (i.e., a missing values issue).
Statistical & end-use models rely on a limited set of fixed parameters. Accuracy of statistical conditional demand models is always challenging because of a regression problem called multicollinearity – that is, explanatory variables that move together and whose effects are difficult to separate in the statistical regressions process. A model that includes both income and dwelling size is an example of this problem. The result is income and dwelling size model coefficients are inflated relative to their true values with biased measures of statistical significance. These problems are difficult to resolve (see “ridge regression and multicollinearity”). Both statistical and end-use models also suffer from the “missing values“ and “missing variables” deficiencies described above.
Advantages of Non-Parametric AI Models Versus Traditional Fixed-Parametric Models
The new age of artificial intelligence has opened an analytic toolbox that offers significant advantages in estimating energy use of household and businesses compared to traditional tools that are still widely in use today.
The great advantage of AI is that it offers “non-parametric” analytics that provides estimation granularity that cannot be provided by traditional “fixed-parametric” methodologies. It does not use fixed general relationships for entire regions or energy customer segments which can miss the impacts of influences that vary within the regions and segments. Instead, it uses information on “neighbors” that share many more characteristics that are not reported but that are reflected energy use and emissions.
About Me: My name is Jerry Jackson. I have consulted on marketing and sales strategies with more than
one hundred companies ranging from
some of the largest energy service and equipment providers to
new technology startups ( see clients).
I have also held professorships at three major
universities and extensively studied and interacted with energy industry customers to develop insights on sales
experiences and customer motivations.
My patented business intelligence software has been
licensed to nearly every major business intelligence software company including Microsoft, Oracle, SAP, SAS, and other companies.
MAISY Energy Apps are designed to provide a similar reach cross the HVAC service and products industry.
I have a Ph.D. in Economics from the University of Florida with a specialty econometrics.
Click Here to Get Notifications of New Blog Entries
|
White Paper Summary: Using Bounded Rationality and Nudge Theories to Motivate Household and Business Energy and Environmental Sustainability at the ZIP-Level: A Public Service Application
April 12, 2022
Bounded rationality is a theory of decision making where individuals apply “satisficing” behavior resulting in an acceptable outcome rather than an optimal outcome. This theory, now widely accepted, was advanced by Herbert Simon in 1956 as a way to explain human decision making in the face of limited or imperfect information, complicated problems, uncertainty and human cognitive limitations. Satisficing decisions criteria are referred to as cognitive heuristics which are mental shortcuts used to make decisions under these circumstances.
Nudge theory was introduced by Richard Thaler and Cass Sunstein in their 2008 book: Nudge – Improving Decisions about Health, Wealth and Happiness. The theory suggests that poor decision outcomes can be improved by acknowledging and using biases, habits and other behavioral and social factors to modify the decision-making process.
This paper presents evidence of poor energy efficiency & sustainability decision-making in the US, discusses the application of both of these concepts to address this issue and provides a framework designed to improve that process. This new framework is used to develop a US ZIP-detailed free online public service App demonstrated with screen shots in a later section of this paper.
The free, public service App can be called as a popup window on any organization’s Web site by
including several lines of script.
Click here
to access the White Paper.
Click here
to access a free (no registration) session.
|
Combating Inflation Headwinds in Home Improvement with a Free Online Energy Cost Saving App
March 22, 2022
Coming off a bumper 2020 and a strong 2021, hardware and home Improvement DIY store sales are likely to moderate in 2022. While rising new home prices and limited housing inventory increase DIY home improvement activity, inflation impacts on DIYers are likely to moderate this trend through the remainder of 2022 and into 2023.
Monthly increases of $300 - $400 in DIYer’s basic living expenses will undoubtedly dampen DIY project demand as nonessential household expenditures are postponed. Increasing DIY materials prices only reinforce this trend.
However, hardware and home improvement companies can moderate these trends by including a free public service App on their Web site that:
-
Provides a 1-minute easy-to-use online App that compares Web visitor energy cost to similar utility customers in their ZIP code – providing a call-to-action for home owners to reduce costs
-
Identifies savings associated with improved efficiency targets and details options for reducing energy costs with low-cost, easy to implement, high-impact actions
-
Expands company online distribution sales channels
-
Increases customer loyalty and
The free App, the Energy Cost & Sustainability Assessment Tool, is offered as a public service and can be called from any organization’s Web site. It operates as a small popup window in front of the organization’s Web page so control is always returned to the organization’s site.
The Apps’ author is Jackson Associates, an energy database and software development firm with 40 years’ experience serving hundreds of US and state governments and some of the largest private sector energy-related organizations. Jackson Associates' business intelligence software analysis patent has been one of the most impactful data visualization innovations licensed by nearly every major business intelligence company world-wide over the last two decades.
Click here
to access a free (no registration) session.
The App is unique in that it:
-
Uses readily available customer data: an estimate of energy costs and a few household/business details to provide an assessment in about 1 minute
-
Provides energy cost and sustainability scores based on a comparison with similar residences/buildings in the ZIP code and presents results in tables and charts
-
Shows reductions in energy costs and emissions achieved for different efficiency targets
-
Provides tips to improve sustainability scores and reduce energy costs
-
Works for all continental US residential and commercial buildings
-
Is powered by the energy industry’s leading utility customer database, the 7+ million record MAISY utility customer database)
The App can be white-labeled and additional pages can be included to provide more detailed marketing/sales information and or messaging.
An Example
A typical family of four living in Orlando (ZIP 32804) with a second refrigerator and a swimming pool in an all-electric home spends $3,606 per year on electricity costs. The Energy Cost & Sustainability Assessment Tool shows that if this household reduces it electricity use to match the best 20% of similar dwelling units in its ZIP code it will save more than $100/month, or $1,288 per year. Most families can match this “best 20%” by following the suggestions shown on the “sustainability/efficiency tips” link with paybacks in a month or two.
Jackson Associates can provide additional App pages to direct online visitors to specific products, learning centers, etc to integrate this resource with other online advertising/messaging.
|
How Clean Is the Electricity You Use?
March 20, 2022
This post was prompted by a question from a user of our Energy Energy Cost & Sustainability Assessment Tool. (https://maisyenergyapps.com/sustainability_tool.htm )
Question: I used the tool to look at emissions for my all-electric home in Arlington VA and got a sustainability score of 41 and emissions of 25,446 pounds. I then changed the zip code to LA where I used to live and got a score of 30 and emissions of 6,709 pounds - a much worse sustainability score but much, much lower emissions. How is that possible?
Answer: Your sustainability score reflects your energy use and emissions compared to similar households in your ZIP code area (59% of households use less energy in your VA ZIP code, while 70% of households in your old neighborhood use less energy).
Your emissions reflect the “cleanliness” of the electric generation system that provides your electricity. Electricity generated in CA results in much less greenhouse gas per kilowatt hour than in VA so your greenhouse gas emissions are much greater in Arlington.
There is good news in these results, though. Each kWh reduction that you achieve in Arlington reduces greenhouse gases by 38% more than the national average so your efforts to improve your energy efficiency will save you money on your electric bill and will have an outsized impact on emissions.
|
Check Your Energy Sustainability Score and Compare Your Energy Costs with the New Energy Sustainability Assessment Tool
January 25, 2022
We have extended one of our MAISY energy Apps to provide a free (no registration required) Energy Cost & Sustainability Assessment Tool that, for the first time, provides individual household and business energy sustainability scores based on ZIP-level comparisons across the US.
Analysis results are derived by applying the widely-used 7+ million household/business MAISY Residential and Commercial Utility Customer Energy Use Databases in a statistical analysis based on ZIP-detailed data.
The Energy Cost & Sustainability Assessment Tool can be freely accessed, with no registration requirement at: http://www.maisyenergyapps.com/sustainability_tool.htm
The Tool can also be launched from any supporting organization’s Web site by including a button to initiate analysis performed on Jackson Associates servers.
The Energy Cost & Sustainability Assessment Tool:
-
Is free to individual households and businesses
-
Uses readily available data: an estimate of energy costs and a few household/business details to provide an assessment in about 1 minute
-
Provides a sustainability score (0 - 100), energy costs and emissions based on a comparison with similar residences/buildings in the ZIP code and presents results in tables and charts
-
Shows reductions in energy costs and emissions achieved for different efficiency targets
-
Provides tips to improve sustainability scores and reduce energy costs
-
Works for all continental US residential and commercial buildings
The tool provides a meaningful call-to-action. For example, a typical household in central Florida with annual electric costs of $3,600 achieves a sustainability score of 48 out of 100. The next step in the evaluation shows that improving its score to match the best 30 % of similar households in the zip code will save $943/year, $4,713 over five years and reduce annual greenhouse gas emissions by 42%. The best 10% would save $1,580/year, $7,898 over five years and reduce annual greenhouse gas emissions by 56%. Emissions savings are translated to meaningful units; for example the 10% efficiency target reduction is presented as emissions savings equivalent to 32,776 vehicle miles driven and carbon savings equivalent to sequestering 13 acres of U.S. forests for one year.
Click here
to see how you score on sustainability and energy costs.
|
Getting Smart About Motivating Utility Customer DR/Efficiency Choices With Benchmarking Widgets
August 31, 2020
NREL (DOE's National Renewable Energy Laboratory) estimates cost-effective residential electric savings potentials of about 20 percent while Department of Energy data show 2018 average residential DR program peak kW savings of 1 percent (for utilities that reported DR savings) and efficiency kWh savings less than ˝ percent (granted some utilities report more savings; however, these industry-wide figures are indicative of savings across the US) .
Why haven’t electric utilities been more effective at reducing system peak demand and electricity use with customer programs?
Clearly customer engagement, motivation and effective calls-to-action to reduce demand and electricity use are being poorly executed by most utilities.
MAISY Energy Apps Utility Customer Benchmarking Widgets are now being offered to overcome these shortcomings by providing utility customer benchmarking tools that provide pop-up windows from individual utility Web sites.
The Widget pop-ups immediately engage customers in exploring demand response and energy savings options. Widgets are easily implemented on utility sites with a few lines of coding that appear as links or buttons.
No utility data are required as the Widgets apply the widely-used Jackson Associates 7+ million MAISY Utility Customer and Energy Use Databases providing a statistically reliable source of ZIP area comparisons. Utility customers input an annual estimate of their electricity costs and limited information on their dwelling units, occupancy and appliance and their electricity costs and carbon emissions are shown as a percentile compared to customers in their ZIP code with similar characteristics.
Customers can identify target percentiles to see how their costs and emissions will be reduced if they meet the best x percentile of similar customers.
At the end of the session, customers are offered links to a series of energy-savings suggestions Web pages, to utility-specified pages or to other destinations.
A more detailed 3-page white paper discusses current energy
benchmarking tools, an evaluation of their shortcomings and why how the MAISY
Energy Apps Utility Customer Benchmarking Widgets provide a significant
advance in achieving DR and efficiency targets by applying past
experience and new research on decision-maker motivational behavior.
The White paper represents collaboration between the Smart Grid Research Consortium and Jackson Associates.
Click here to download
the PDF White Paper:
New Online Benchmarking Widget Motivates Utility Customer “Smart” Decisions
|
Are You Passing Your HVAC/MEP Web Visitors on to Your Competitors?
July 22, 2019
Web marketing studies show that visitors spend an average of 8 seconds before moving to another site. The average contact conversion
rate for our industry is just a bit more than 2% meaning that just 2% of visitors actually follow through with a company contact.
Actual sales conversion rates are substantially less, so if your Web sales conversions are disappointing, join the club.
The answer, of course, is to provide an inducement to your Web visitors to stop and engage with your Web site, company and services.
So, how do you stop that potential customer’s search process to focus on your company?
If you look at your and your competitor’s Web sites, you will find that while details may differ, few Web sites offer a really
compelling reason to stop and engage as opposed to moving on to see if another company has a more convincing sales argument.
It’s instructive to look to other industries to identify more effective customer engagement strategies. Financial firms have
successfully used mortgage/loan payment widgets to engage potential clients by providing answers to questions
such as “how much can I save on my mortgage payments if I refinance now.” Once a customer engages in this kind of interactive
process, conversion rates soar.
MAISY Energy Widget popups (which draw data from nation-wide MAISY Utility Customer Databases) now provide this same customer
engagement opportunity for HVAC/MEP and other energy service companies by presenting and answering the following questions:
-
"How do my energy costs compare to peer buildings in my area?"
-
"How much can I save in energy costs if I improve my ranking to be among the best 10th, 20th, 30th or 50th percentiles?"
-
"How would these improvements impact my sustainability footprint?"
Offering answers to these questions with an easy to use pop-up on your Web page is an effective call to action to stop the
search and engage with your HVAC/MEP firm.
|
HVAC Companies Have a “Commodity-Service Sales Problem”
August 12, 2019
This note dives a little deeper into difficulties encountered in marketing HVAC services/products on the Web and how to
address them. Most HVAC company Web sites advertise generally similar products and services resulting in
a “commodity-service sales problem.”
From a customer perspective HVAC equipment offerings resemble a commodity – that is a product that is similar or interchangeable with products provided by competitors. Consequently HVAC companies try to differentiate their services to “stand out.” However, a quick Google search of HVAC sites shows that most HVAC Web sites offer very similar services and sales propositions.
If your Web visitors move on to the next site and see a similar sales proposition, they are likely to continue on to the next site, and if that one is similar, continue on and and so on. They may potentially come back and contact you but, the probability of engaging with you drops dramatically as the search continues. This behavior explains the average 8 second look-see and dismal conversion rate for most HVAC-related sites.
The answer, of course, is to attempt to match features of your competitor’s Web sites and to differentiate your Web product/services offerings from those of your competitors in a way that either stops the search on your site or stands out in the mind of the searcher sufficiently to result in a return to your site.
To assess and address your Web site’s comparative sales proposition advantages/ disadvantages, do a search on a
phrase that you expect your customers to use – for example “commercial HVAC service.” Review at least the
top 20 sites to see what Web features your competitors are using in an attempt to differentiate themselves. Then:
-
Make sure that your site includes all the common, easy-to-replicate features that your competitors use. For example,
a 24-hour emergency number is an easy addition. If your competitors highlight their warranty, do the same for yours. A $50 first-time coupon for home owners is another easy feature to include. Your objective is to negate the differentiating features on your competitor’s sites. If you don’t have testimonials, be sure to add that feature.
-
Next, identify competitor’s more unique site feature that might make a visitor stop or that really stand out. Evaluate the costs and benefits of adding these features and update your site with features you feel are appropriate and likely be valued by Web visitors. For example, adding a page that describes energy star benefits and provides links to benefits of energy star buildings would require a relatively easy addition to your site. Adding a live chat app is a great idea if you have a sales person who will always be available to interact with Web visitors on at least a reasonably technical level.
-
At this point, you will have, to the extent possible, minimized your Web site sales proposition disadvantages relative to
your Web competition. Now take the extra step of identifying an “eye-catching” feature that does not exist on competitor sites that you feel can provide a significant differentiation for your business. Is there something about your products or services that will make you stand out from competitors in your market area? Conduct searches in other markets to see if you can adopt a new Web feature that is not common in your market. Consider third-party add-ons like the Energy Sales Widget on your Web site.
This note focused on how to limit the Web differentiation effectiveness of your competitors with a final recommendation
to add something significantly different and memorable to your Web site to stand out from your competitors.
A forthcoming note will discuss the value of gaining a better understanding of your Web customer market.
|
