In order to do that, a business needs to do the following:
- Gather secondary research
- Gather primary research
- Analyze the research and turn it into an actionable plan (that will be part 3 of this 3-part series)
In addition to the research we do for the articles we write about AI across industry sectors, we have to do this kind of research for clients. If we’re going to explore the process of garnering the go-to-market insights that we need, we might as well use an example.
Let’s say that there’s a company that’s looking to sell conversational interfaces into the banking space. The first step is to gather secondary research. The company won’t do any direct interviews with people but instead finding the insights in the world that can help them frame the questions they want to ask of real experts, real buyers, real vendors.
Step 1: Secondary Research
Market Sizing and Growth Rates
In the secondary research phase, a company is often going to be looking first at market size and growth rate. If they’re going to go to market, it makes sense to figure out if that market fits their criteria. We talked about this in part one of the series:
- Is it a large enough market to be worth it?
- Is it a market that seems to be growing or shrinking over time?
Obviously, the vast majority of the time a business is going to be looking for markets that are growing because they want to be a part of that growth. That said, there probably isn’t any market sizing or growth rate reports out there for conversational interfaces in banking.
What a business can do, however, is find the potential existing market sizing and growth rate research that might be available for free online that could again inform sort of what we’re looking to decide on even if it doesn’t have to do with AI. They could look at proxies that might be the market size of customer service platforms as they’re used by banks and financial services companies. They could look at internal communications platforms and what the spend is on those kinds of platforms currently and in terms of growth rate for banks. They might look at other kinds of marketing communications and messaging technology that banks are using.
Oftentimes, this kind of information is going to be found by just Googling. Here at Emerj, we use proprietary data sources that we’re forced to pay for every year, but not everybody is going to be using those kinds of resources. You might just be able to hit the web.
Generally speaking, if you’re going to hit the web you’re going to run into a lot of less-than-reputable market research sources. You’re going to find various firms who will have a landing page that ranks well for a term and some numbers up on that page for the market size of a specific sector product. These numbers are often not trustworthy. A lot of the time there hasn’t been any primary research behind that kind of information. These market research firms have a particularly shotty reputation amongst folks in the market research world in and of itself.
Ideally what we’re looking for when we look for market sizing and growth rates is reports from industry associations. If these associations are publishing numbers about the growth or the size of a sector, I trust these numbers to be much more likely to be true for two very important reasons.
An industry association has absolutely no incentive to fake a bunch of numbers to get you to find them with a search engine. A market research firm somewhere might be able to just flub some numbers that they lay out for free in order to have you come over to them, and that might benefit them in some way. An industry association really is only going to lose face. There’s really no upside to that.
Also, an industry association reaches a lot of people in their sector. They’ve already got a lot of bankers or retail folks, so they’re garnering insights from the right people.
Business can also look at their competitors and existing applications in the market. They can find out what are the relative market shares for their competitors. With regards to AI products, it’s unlikely that this information is going to be found easily if it exists at all.
What a business might find instead is five or six different companies that are potential competitors, and they might see some that have raised $100 million dollars and some that have raised $9 million dollars. They might see some that have Wells Fargo and Bank of America on their website with testimonials, and they might find some that have absolutely no testimonials and no big bank logos on their website.
Businesses might have to again proxy an understanding of market share. That’s often a decent starting point. They could do the same thing for solutions at banks that aren’t using AI. It might be much easier to look at customer service software within banking. They might just look at who’s offering customer service software in banking and what their relative market share slices are. That would be something they could find in a report and they could figure out what it is about the companies that are winning market share that is helping them win?
Categorizing the Vendor Landscape
Getting an understanding of that might be useful if a business is looking to go into the same space with AI. If they’re looking at the actual artificial intelligence vendors in this space, a business might want to potentially score them across different kinds of criteria.
At Emerj, for example, if we’re looking at the vendor landscape in any given domain, whether it’s pharma or banking, we’re looking at specific scorecards for different applications based on proxies for ROI evidence. Essentially, which of these companies seems to be showing signs that they’re actually delivering results for clients, not just running pilots and bragging about themselves online?
A second way we score vendors is on level of adoption. For example, how many banks are using a conversation interface?
A third criterion is traction, which entails the number of case studies that are vailable across the conversational interfaces in banking space. Banks might be adopting conversational interfaces, but can a user go onto that bank’s website and use it? That matters.
Step 2: Primary Research
Once we have a sense of our secondary data, it’s time to hit the phones. It’s going to be time to send the surveys. It’s going to be time to collect primary research. This is the hard part.
The hardest part of primary research is reaching buyers. At the end of the day, the people you ultimately want to garner your insights from are the people who are going to be spending money or the people who are working directly with the people who are going to be spending money.
Oftentimes a business is going to have to offer them something for getting on the phone with them, such as an executive summary of the completed research that they could check out.
Here at Emerj, we have a little bit of a shortcut because we have tens of thousands of business leaders with an AI focus in very important sectors on the website every week. These people are on our email list. If we want to garner insights from marketing leaders within retail banks, those are people we can contact pretty quickly because we have a database and know who reads what and what they’re up to.
That’s not necessarily something that every vendor company or even every enterprise is going to have at their fingertips, but it is still something that people could do by having an intern scrape LinkedIn or hiring a firm to help with this sort of work. But finding buyers is really the biggest deal because it provides a sense of who it is that buys the product that a company wants to sell.
Initially, you’re going to be taking guesses. A business might think or four company roles that they might want to reach out, and they’re going to want to reach out to all of them.
What you want to ask those folks are questions that are going to get at the pivotal insights we talked about in part one of this series, namely double checking with them who buys the product. This breaks down to a couple of things:
- Who controls the budget?
- Who cuts the check?
- Who is part of the buying decision?
There’s a lot of different facets to that question, but that question is really primary. If you can get frank answers to that, you are more than half of your way to the kinds of insights you need, which is namely those from buyers.
When you’re speaking with buyers, those other questions can be more or less summed up thus. Businesses will really want to get a sense of what buyers are running towards and what they are running from. For example, what are the opportunities that they’re chasing? What are the things that they would do anything to get to, that would better their circumstances, that would better their workflows, better their earning potential? What is motivating them forward the most?
If they’re looking to adopt conversational interfaces in banking, are there opportunities for efficiencies that they’re excited about? Are there opportunities for consistency in terms of how customers are treated that they’re excited about? Are there elements of cost savings or elements of the hiring process that they’re excited? What are they driving towards that they’re very excited to reach? Essentially, what are their positive motivators?
As for what they’re running away from, what are they afraid of? Are they afraid of being left behind in a new modern era of customer service? Are they afraid of having astronomically higher costs in their call centers than their competitors do? Are they afraid of letting a machine take control of the customer experience?
If a business has a sense of that, that’s ammunition that they can use for selling their product, for appealing to a real need and making sure they’re offering a solution that can actually be valuable.
Existing Non-AI Vendors
In addition to that, they’ll want to ask questions about who buyers trust and are working with now. In the conversational interface world, this might simply be asking them what kind of solutions they’re paying for in terms of existing customer service, call center, or email software. It might have nothing to do with AI. They might not be spending with any AI vendor.
Getting a sense of sort of what the existing solutions are that buyers like or don’t like is critically important because the business can get a lay of the land of, again, how to fit their product into the buyer’s workflow, which is something that AI products often do a very poor job with.
Thinking about the way to integrate your product into existing workflows is the most important part of the strategy. This is absolutely pivotal. It is often overlooked, and it is one of the critical reasons why AI has such a dismal current traction rate in the enterprise today. So, getting these insights is something businesses want to do earlier rather than later.
Another question they might ask is if they are working with or talking with any existing AI vendors. If they are working with them, businesses will of course want to understand who they’re doing business with and what their experience has been.
The second group that they’ll talk to when it comes to primary research is other vendors. Again, if a company is looking at conversational interfaces in banking, they could talk to vendors that are selling conversational interfaces into the banking sector with artificial intelligence embedded, but they might also want to talk to the vendors that are currently being used at banks.
In large part, asking vendors allows a company to get another perspective on the same core insights that they’re looking for. Common questions that will be very high on the list are the following:
- Who is the buyer? And who else is involved in the buying decision?
- What has been the most helpful way to get in touch with buyers?
- What has been the most helpful way to get buyers to buy?
- What areas of the market seem to be spending the most and have the biggest need?
Though individual projects might vary, these are going to be the critical questions in go-to-market strategy.
Step 3: Analysis
In the analysis step, businesses are going to do a much deeper dive into the insights they’ve garnered. Of course, the more insights the better. If they are tasked with doing a research project, they’re always going to have to have enough interviews with buyers and with vendors to actually be able to show charts and graphs.
If a business is doing something for the C-Suite, they’re going to want something simple to look at to understand the insights. They’re not going to want to read the Excel sheets. If they’re doing a very small project within a specific department, then perhps they don’t have to visualize things.
But if they’re doing a bigger project and are going to be spending millions to launch something into a sector, often they are going to need to get enough insights to visualize them, to create the charts, create the graphs, and get an understanding of who thinks what parts of customer service are growing, for example.
The more the better. It’s not always possible to get a full 30 buyers or 20 vendors in a survey, but they’re going to want to get close to that if they want insights that they can actually put money behind. So analysis is going to be drastically different from project to project depending on what a business wants to learn and how they want to learn it.
Turning analysis into insight requires asking questions about the data. For example:
- Is there a reason that the vendors had different opinions about how to build trust with buyers than the channel partners did?
- Are there reasons that they disagreed so sternly?
- Should they trust one of them more than the other?
- Are vendors likely to be misleading us because they don’t want to admit something?
- Are channel partners misconstruing some bit of insight?
- Why is it that these two answers are so different?
Analysis might make a team realize that there are must-have insights that still aren’t clear from the results of the research, and more surveys or secondard research is necessary before making a decision with company budget.
Maybe even after talking to buyers, they realize that maybe only three or four of the buyers they talked to were really the true economic buyer, and they don’t have a good enough sense of the motivations of what they’re moving towards and moving away from for those three or four.
Poking into that is going to be critical for coming up with the kinds of insights that a business can put dollars behind.
Oftentimes at the end of the analysis, they’ll find that they might want to do another six phone calls, another 10 phone calls, to get to the root of what it is that they’re actually after in the first place.
Often from there, they’re going to come up with an action plan and a map that can put these ideas into action, and that is going to be the third episode in this series on go-to-market for an AI product.
In the next article in this series, I’m going to be exploring what it looks like to take secondary research, primary research, and analysis and to turn it into an actionable plan that an enterprise or a small company could use to hire more effective employees, to revamp their marketing strategy, and to be more effective and garner the attention of buyers.