By Abby Yolda January 2, 2024
Overreliance on artificial intelligence tools and chatbots like ChatGPT, Bard, Dall-E and others is a losing game. But that doesn’t mean solar companies should turn their backs on this new technology. Solar companies are all about embracing emerging tech, right?
Energy Circle, the digital marketing company where I work, focuses exclusively on contractors, businesses and organizations in the clean energy and energy efficiency sectors. We’ve been experimenting heavily with AI tools since 2022.
We’ve always operated under the guiding principle that great marketing requires real-world expertise and a human touch. Adding AI into our toolbox doesn’t change that. Any solar company curious about artificial intelligence should know that there are plenty of ways to benefit, but only if you strike the right balance between human and robot.
The Benefits of Using AI in Solar Marketing
AI Tools Save Time
It’s a common scenario for solar contractors: they’re already strapped for time, knowing that they should be out on a job site or helping solve business problems, but they’re also trying to carve out time for the marketing that they know they’ll wish was done six months from now.
AI can do a lot of different tasks — and it can do them quickly. Think about what it would be like to do all this marketing 15% faster. It might only save 10 or 15 minutes with each task, but those things add up quickly, and a 15% improvement is a conservative estimate.
For example, there are AI tools out there that could help a company design a new website, write all of the content for the site, and plan out a marketing strategy for the next 12 months. The team could do all of this in an afternoon.
Here are some examples of actual marketing work that AI could help a solar contractor do:
- Making the Sales Cycle More Efficient
AI can help an existing sales team do more, creating efficiencies within the sales process, supporting design, helping to draft proposals, and responding to customer reviews. AI apps can even record sales meetings, creating transcripts and summarizing key points for future reference. These transcripts aren’t always 100% accurate, but the technology is constantly improving and now they are usable for many use cases.
- Creating Content
AI can write copy for blogs or webpages, draft emails, create custom images and videos, design infographics, write up social media posts, and generally help solar companies brainstorm and draft all different kinds of solar content.
- Analyzing Sales and Marketing Data
AI is really good at analyzing lots of data. It can help forecast sales, take audience lists and segment them into different groups for marketing or outreach campaigns, and much, much more.
This all sounds pretty abstract, so let’s give a few actual examples of how this could work in action.
Example #1: Segmenting historical customer data to build new marketing campaigns
Let’s say a sales team has the idea to create a new marketing campaign targeting past customers who installed solar panels but not solar batteries. The staff want to reach out to only those customers via email and a physical mailer, but they don’t have a sophisticated CRM platform like Salesforce — all they have is a spreadsheet with columns of customer data.
Having to manually parse through that spreadsheet could take a long time. But AI tools like ChatGPT can make data analysis surprisingly fast. Through the use of the Code Interpreter plug-in (or third-party plug-ins like Noteable, which are designed to do similar types of data analysis), the team can upload that spreadsheet and quickly get segmented lists of email and mailing addresses to start campaigns.
Example #2: Forecasting month-by-month sales for 2024
Maybe a manager is sitting down to review their company’s performance for 2023. Again, they have lots of data detailing sales that go back years, but how can they use this data to better prepare for next year?
Using the same data-analysis tools that we mentioned in Example #1, they can input their historical sales data and ask ChatGPT to quickly help them find sales trends or make predictions about sales in 2024.
ChatGPT can analyze this information. And it can also take that raw data from spreadsheets and create customizable charts and graphs to use in future sales presentations to the company. And it can do all of this in a fraction of the time that it would take a human.
AI Gives Solar Companies a Competitive Edge
Smaller solar companies have often struggled to stay competitive with the solar giants that have more resources and larger marketing budgets that they use to stay ahead of their competitors. AI can level that playing field for smaller, local installers and up-and-coming contractors.
But the benefits aren’t just for underdogs. If a bigger company already has a healthy marketing budget, AI tools can help the staff do even more with it.
The Risks of Using AI in Solar Marketing
We’ve been painting a pretty rosy picture so far. But before companies drop everything and convert 100% to AI, they should be aware there’s a less-enticing side to this technology as well.
AI Is Good — But Not Great — At What It Does
Remember when I told you just how much AI tools could do in a single afternoon? Well, I didn’t say anything about the quality of that work. ChatGPT might give a company a starting point, but it’s not going to provide the kind of unique and emotionally intelligent content the team needs to connect with homeowners and differentiate itself from its competitors.
Solar is a big-budget purchase — and homeowners don’t choose contractors lightly when making that kind of investment. The content needs to demonstrate that the contractor knows what they’re talking about and understands customers’ pain points.
AI can confidently give completely false information — in the information technology industry, this phenomenon is called hallucinations.1 And it does so in a way that makes one think it knows what it’s talking about, so unless you already know the information is wrong or you fact-check, it’s easy to be fooled. AI tools have a surprisingly hard time telling you, “Sorry, I don’t know.”
Just one example: This morning, I asked ChatGPT about the solar tax credit and whether I was eligible for this incentive. But ChatGPT is only trained on data up to January 2022, so it didn’t know (and couldn’t tell me) that the Inflation Reduction Act overhauled and extended the tax credit in 2022.2
The AI Revolution Is Going to Usher in a New Era of Content Marketing (and Raise the Bar)
As more and more AI writing is published, the bar for “standout” content will be raised.
Google uses E-E-A-T (Experience, Expertise, Authoritativeness and Trustworthiness) guidelines to refine its ranking algorithms.3 Essentially, the more content meets these guidelines, the more visible it is likely to be in search results.
Google has gone on the record to say that the (“appropriate”) use of AI-generated content is not against its guidelines.4 But that doesn’t mean AI-generated content is automatically going to rank. AI tools don’t always follow Search Engine Optimization (SEO) best practices and, given the rate at which SEO best practices change, may not be able to keep up with them.
Plus, AI content itself often falls somewhere between “okay” and “good,” not “great,” and that type of content is unlikely to rank well regardless of who has written it.
Striking a Balance: Making AI Work Better with Human Elements
Start with AI, Don’t End There
It’s a good analogy to think of AI like a competent personal assistant. There are many tasks for which a manager might use an assistant, but odds are they aren’t entrusting them with important work without signing off on it or giving notes first. AI is the same way.
- If you use AI to generate copy, edit and revise it to make sure it meets the same standards you would hold for your own writing.
- If you use AI to analyze data, make sure you don’t stop all analysis yourself. AI might come up with something that wouldn’t have occurred to you (or arrive at the analysis more quickly), but that doesn’t mean you don’t have takeaways that AI couldn’t develop.
- If you use AI to brainstorm ideas or marketing strategies, see what you can add to or change in the results that take them from passable to exceptional.
Fact-Check, Fact-Check, Fact-Check
If you don’t know enough to verify the accuracy of a piece of writing from AI, you need to fact-check it, no matter how big or small a claim.
This is especially true in the solar industry, where regionality and policy play a huge role.
For example, ChatGPT often touts the benefits of net metering when asked to write about the benefits of solar. But, as any solar contractor knows, net metering isn’t available everywhere and it works differently in different places. If companies don’t edit their content to include specific information about their regions, they’re not providing information that’s relevant to their customers (or true).
Companies Should Make Sure They’re Telling Their Story About Their Businesses, Speaking to Their Customers
AI tools know a lot, but here are just a handful of things we’ve seen through experience that AI is not very good at talking about:
- Any kind of writing that depends largely on current information, including:
- Energy incentives (like tax credits, solar renewable energy credits, modified accelerated cost recovery system tax depreciation, and other local rebates and incentive programs that may be subject to changes)
- The latest products and tech specifications from specific manufacturers
- Solar policies (like updates to net metering in your state; for example, ChatGPT doesn’t know about Net Metering 3.0 in California)
- Brand voice (Is a company a solar tech expert? Is it the go-to company for solar + storage? Is it a veteran-owned business? Is it a B-corp with clear mission and value statements? It’s difficult for AI to authentically capture the way a company “speaks” online without extensive training and prompting.)
- A company’s history (Has it completed more installations than any other local solar company? Does it have lots of specific experience with off-grid solar or solar carports? Is a solar installer a local celebrity? Again, without lots of training, AI tools have no way of knowing what sets a company apart from any other solar installer.)
- The unique characteristics of a company’s customers and its service area (How does the seasonal weather and geography in the service area affect the types of solar arrays the team installs or how they install them? The marketing messaging will likely be very different if homeowners in the area tend to be more environmentally motivated than savings-focused or vice versa.)
We use the output of ChatGPT and other AI tools as a framework to build on. Then we make revisions — lots of them — as we think of additional ideas. The final product is always dramatically different (and better). We save some time in the early stages of the process, but never sacrifice our voice or cut corners on quality.
In our experience, this is the best way companies can integrate artificial intelligence into their marketing work and create results that are customized to their businesses and, most importantly, effective.
Unlike the sharp rises (and equally sharp falls) of digital assets like NFTs5 or the trendy, invite-only social audio app Clubhouse,6 AI technology is here to stay, and it’s only going to become more and more essential to any future solar marketing work.
That doesn’t mean that every solar company that embraces AI will see success. But the contractors and organizations that find the right mix of human and machine, leaning into the strengths of each component, will be the ones to truly capitalize on this rapidly advancing new technology.
About the Author
Abby Yolda is the director of solar and digital marketing strategy at Energy Circle. With a background in public policy and digital marketing, she has over 15 years of experience working with contractors, businesses, nonprofits and trade associations across the solar and energy efficiency sectors.