When Google Falls Short: Smarter Research for CRE Pros Using Generative AI

Whether it’s a project, a client’s specialty, or lead prospecting, research is something that’s part of every CRE professional’s life. We all have our preferences for sites and sources we use, but for years, the gold standard for web search has been the Google search engine. Google is simple to use, as your search queries deliver links to websites that contain the information or sources you need. 

Until recently, Google and other search engines have shown primarily the road to the information you need, not the information itself. Now, the line between “search” and “answer” is blurring with the introduction of Google’s AI Overviews and Gemini-powered summaries that not only deliver answers within search results but also link to the sources.¹ 

And so what about for complex research requirements? Or time-sensitive research and relationship-driven decision making? 

Generative AI like ChatGPT, Claude, and Gemini fills some of the gaps that traditional search can’t. AI tools work as digital personal assistants: They answer your questions, provide sources for information, and some even have built-in browsing abilities that enable them to analyze websites and compile the most important information. AI is emerging as a useful tool for market research, content creation, and even business development in CRE. Related: 8 AI Tools for the Real Estate Market

What Traditional Search Does & Doesn’t Do Well

What traditional search does well

At this point, traditional search has been around for decades. Much of the internet’s architecture has been structured around organic search, from keywords to headlines and metadata. And so traditional search is still a useful tool for research! Some of the positives of traditional search are:

• Providing real-time news and updates 

• Direct sources for information

• Website indexing (such as finding a company homepage, contact page, and other specific information)

• Accessing peer-reviewed research (through services such as Google Scholar)

• Broad industry reports and data

Where Google falls short for CRE

But in some spots, traditional search can fall short in the depth of complexity of the research. Some places where traditional search struggles are:

• Understanding the context of your research (e.g, "Who should I talk to at a mid-size multifamily developer expanding in Phoenix?")

• Succinctly summarizing the key information from a source

• Synthesizing multiple sources into useful takeaways

• Creating original, custom content

• Surfacing hidden insights (e.g., how a materials supplier might tailor their pitch based on a developer’s recent deals)

What Generative AI Does & Doesn’t Do Well

What generative AI does well

Despite generative AI being a newer tool, it’s still proved useful for automating tasks and saving time in research. Generative AI doesn’t only answer questions! It can scan documents and websites for key information that are pertinent to the context of your business. Some of the things that these language learning models excel at include:

• Conversational answers to natural language interactions (for example: “Summarize the last 3 deals for this company and suggest how a lender might approach them.”)

• Using retrieval-augmented generation (RAG) to comb through research papers, case studies, and websites for key insights

• Researching specific job roles and decision-makers²

• Analyzing patterns in CRE across multiple sources (e.g., leasing, building trends, or product pricing)

• Drafting original content for marketing, web, and even professional email communication

Where AI falls short for CRE

However, generative AI isn’t perfect and is still early in its development cycle. Here are some areas where AI can fall short:

• Sometimes provides inaccurate or out-of-date information that needs to be verified or clarified (requiring multiple rounds of querying the AI)³

• AI doesn’t always understand the full context of your business or business goals

• Requires a time investment for initial setup to “train” the AI on CRE and your business

• Scalability and cost for subscription plans 

How Generative AI and Search Work Together

Our suggestion is a dual approach: Use both traditional search and generative AI together for a more efficient process. Broadly speaking, take advantage of the strengths of both technologies while avoiding the pitfalls that may slow down your research. Traditional search adds discovery and breadth, while generative AI offers (a mostly accurate) contextual analysis.

For example, Google and other search engines can locate useful, timely sources for your research. You can source news, white pages, and websites that are relevant to your project. Narrow your sources by using variations on keywords and search terms. 

Once you’ve compiled your sources, explain to generative AI what the goals of your research are. From there, provide it with the sources that you’ve collected and instruct it to analyze and summarize the most important data within the context of your research. Be sure to verify the information that the AI has compiled and explain to it what it may have missed or gotten wrong.

Some quick tips for writing prompts for AI are:

• Be as specific as possible with your prompts. Clarify what the AI gets wrong or seemingly doesn’t understand.

• You can feed AI websites, PDFs, and other third-party links for analysis — no need to copy and paste text!

• Ask for follow-ups on content. Refine what’s been generated for you. Treat your AI tool as an assistant.

• Give it data that has a clear structure (e.g., job titles, company names, city names).

• Tell the AI to “Act as if…” they are a specific role or company so it can emulate a response around your CRE goals.

• Provide the AI with things not to do (e.g., “Ignore all businesses outside a 50-mile radius.”).

Once the context and most important information are understood by the AI, it can help you write, analyze and even plan around the data that you’ve found. If your research suggests there’s an increase in mixed-use development investment in your city, for example, AI can may assist you in identifying and developing messaging to contact a developer for a potential business partnership.

How Biscred Leverages AI and Traditional Search

Those familiar with the Biscred platform already know its ability to perform conventional database queries using filters and keyword search terms. With over half a million companies and 4.4 million professionals combined with numerous ways to slice and dice queries, Biscred has become the most robust prospecting platform in the CRE industry. 

We’ve added an AI-powered query builder to our platform, offering the best of both worlds — traditional search and an LLM-powered query builder to help you find what you’re looking for quickly and efficiently. 

Here’s how it works. 

Step 1: Select the AI Powered Query Builder 

Step 2: Enter your query using natural language.

Step 3: Biscred builds your list.

The simpler the query, the quicker your results.  

Step 4: View and refine your results.

In this example, we found 760 companies and 6,605 professionals that matched the query. From here, you can use traditional search filters to further narrow your list based on industry, location, company size, property count and dozens of other filters.

Schedule a Demo

Quick Answers to AI vs Google Questions

Is AI a search engine? Is ChatGPT a search engine?

AI has similar functionalities to a search engine but is not technically a search engine. The main difference is how an AI like ChatGPT can generate new content. Some AI tools can access the web to access information based on the questions you ask, while others rely on pre-trained data. However, AI will attempt to analyze and compile this information rather than just provide a list of results. 

Simply put, the main difference between AI and search engines is its ability to generate information, analysis, and content. 

Do search engines use AI?

Yes, it’s becoming increasingly common for search engines to integrate machine learning or an AI tool into their functions. Google Gemini (Google’s AI) is directly integrated into the search engine, even having an “AI Mode” that uses real-time search to generate AI summaries based on Google’s indexed content. . 

On the backend, search engines will sometimes use machine learning to tailor better results for your searches. Machine learning can help understand misspelling, provide personalized results, and even disregard sources not as relevant to what you’re searching. 

Is AI better than Google?

It depends on the task and situation. AI is better when analysis and content generation are needed. Google is excellent at sourcing new information.

A simple search for “real estate investors” on Google will take less than a second to give you results of local investment firms, having already considered your location. An AI tool may be able to define real estate investors, while some may require additional context or live data access to interpret your web query.

However, if you want to generate email copy for contacting local real estate investors for potential business partnerships, generative AI is far better at providing content based on the needs and concerns of your business. You can even have generative AI take into account the specific goals of an investor based on their website. 

🔍 Comparison: Google Search vs. Generative AI in Commercial Real Estate Workflows

NOTES

¹As of mid-July 2025, in-search AI tools such as AI Overviews and Bing Copilot can be hit and miss with their accuracy. Verify everything these tools deliver in search engine results pages. 

²AI can infer roles from org charts, LinkedIn profiles, and company bios, although the accuracy depends on available data. Biscred’s database augments machine learning with data analysis (performed by real people) to build the biggest, most robust CRE prospecting tool in the industry.

³ AI models, especially LLM (large language models), are known to generate false, misleading and nonsensical information that appears to be accurate or even coherent. Always validate the information any AI model delivers. 

Photo credit ID 346220822 | Ai | Rafael Henrique | Dreamstime.com