I'm not a Luddite, one of those 19th-century factory workers whodestroyed machines because they thought industrialization wouldreplace their jobs. On the contrary, I love new technology.

In the mid-'80s, I created what has become one of the largestand most successful self-insurance pools in the nation forpolitical subdivisions. Part of this process was a feasibilitystudy: That study occurred prior to the proliferation of Lotus1-2-3 (predecessor to Excel) so I did all of my projections with anadding machine along with yards and yards of paper tape. 

We used two-dozen Apple IIGS computers to leverage ouremployees' abilities to could keep the expense ratio for thatinsurance company under 27.5%—which compared favorably to thethen-common expense ratio of about 35%. Using Appleworks andspreadsheets that I devised, my employees could quote a premium forthe average city in under 30 minutes, something that would havetaken them five to seven hours using paper and calculators.

In short, I love computers. And, like most, I spend a large partof every day on the Internet learning and communicating with myclients and companies. Even so, I think most of what we read andhear about the success of agents using their individual websites toattract a large volume of new business is a hoax.

The promise (in theory) …

For the last four years, I've been involved in creating a largewebsite meant to increase the efficiency of Internet marketing forindependent agents. So far our efforts have been one long, veryexpensive lesson. We've invested a significant amount of money andtime and have developed a system that works, but not yet on a scalethat would satisfy a large paying customer.

The world of websites is highly speculative, as Google sets therules for ranking on a search-return page and won't tell anyoneexactly what those rules are. The search giant's fickle andsecretive nature has added to our cost and has increased the widthof the moat for those who want to develop cost-efficient leadsthrough websites.

A few days ago, I attended an all-day seminar held by one of ouragency's contracted insurance carriers. The instructors stated thatif insurance agents:

1. created a visually pleasing website,

2. posted interesting content,

3. included relevant captioned pictures,

4. paid attention to keywords,

5. properly titled their pages,

6. made sure their site was responsive to mobile devicesand secure(https),

7. claimed the proper local search pages,

8. blogged frequently,

9. wrote interesting meta descriptions,

10. made sure they had positive reviews from clients,

11. responded properly to those reviews, and,

12. became proficient in website analytics,

… then they would have leads pouring into their agency.

Utter nonsense.

Unfortunately, the average agent who followed this advice wouldbe out of business. They'd have no time (and limited marketingresources) to sell or service their customers, because maintaininga good website and attracting page visitors requires a significantinvestment of time and money.

Google does offer advice. One of Google'sprimary tenets is, "Make pages primarily for users, not forsearch engines." That seems to oppose the primary purpose of theseminar instructors' list. Similarly, most of what they suggestedappears dated and ineffective when compared to current expertopinions.

… and the sobering reality

Think of the Internet as a very large convention. Your booth isone of approximately 500,000 other booths competing for attention. In order to "rank" high enough so that Google places youon the front page (or in the main "convention aisle"), you need tohave content that is unique and desirable. Then you need to have anInternet pedigree, which very, very few insurance agencies canafford.

An Internet pedigree is often defined as your Domain Authority(DA). It is a number between zero and one hundred, with a onehundred being the best. Most agencies have a DA in the teens. Mostinsurance companies have a DA of between fifty and eighty. DA isbelieved to part of the algorithm used by Google. Many think it iscombined with Page Rank and several other factors to help Googledecide how to rank your page when they respond to a search.

For example, when "auto insurance" is the search term Googletells me they found 156,000,000 results. The top results areinsurance companies with high DAs. Even the mega insuranceinformation website, Trustedchoice.com—which has a high DA—came inon the fourth page on the day of my search. The first independentagent listed was on the seventh page. Almost no one goes beyond thefirst page, so the worth of coming in on the seventh page isnegligible.

A search on "auto insurance" is so competitive that the cost forPay Per Click for that term is over $73. That is a lot to pay forone website session. Even with a conversion rate of 20%, which isaggressive, that would put your acquisition cost at over $350 forjust that portion of the expense.

Long-tail search

The instructors in our seminar suggested that we concentrate onlong-tail search. "Long-tail" are the longer and more complexphrases people use to search. It is widely held that themore esoteric the search, the easier it is to rank on the firstpage. That is excellent advice. Except in this very competitiveworld it is becoming harder and harder to rank, even forlong-tail.

Our site haswell over 1,000 indexed pages. Indexed pages are recognized byGoogle. Yet we struggle to have about 275 keywords ranked on thefirst two results pages.

For example, if you search for "what do I do when my car hasbeen keyed" our site ranks second. We don't have "keyed" as akeyword. In fact, we don't pay all that much attention to keywordsanymore. Over a year ago, Google announced that more than one-fifth of all searchesentered in the search bar had never been used before. People arebecoming more and more specific in their searches, and are writinglonger inquiries.   

Google's bots can read for "context," so keywords seemingly havediminished in importance. The bots will match inquiries with whatthey think is the most relevant, unique content. Supposedly ifGoogle finds two pieces of content that are the same on twodifferent sites, they may considerit a deceptive practice and an attempt to manipulate searchengine results. Google does not like to be "scammed," and willpunish with low rankings those websites that attempt to put oneover on them.

Another of my contracted carriers has a large Internet effortthat helps agencies with their sites. I've used analytical tools onthose agents' websites who use their services. Those tools allow meto gauge the website's traffic, view their keywords and dozens ofother pieces of relevant data going back for several years. I'venot been able to see any difference in traffic levels for themajority of the sites I've checked before and after those agenciesspent hundreds of dollars a month for this service.

Further, we've looked at individual sites and have found thatthis "service" posted identical content on several sites. As statedabove, this is considered very bad practice by most experts in thesearch engine optimization field.

Well-intentioned advice

A few years ago I attended the annual meeting of one my agency'scontracted insurance companies, during which the company assemblesthe top 1% of its agencies. They had an owner of a local agencyspeak to us in a breakout session about their agency's website.According to the company and the agency, they had done everythingright with their site. It was obvious that they had spent a greatdeal of time and money on it. Their site currently has a DA of 41,which is one of the best I've seen for an independent insuranceagency.

I recently checked their site with my tools, and they're gettingabout 3,000 visitors a month. Given standard conversion ratios,that would mean they're getting between 10 and 20 inquiries amonth. That amount of activity could be purchased from leadscompanies for far less.

A friend of mine has had a good agency website for years. It hasa DA of 16, which is within the average range for agencies withsites. He has contracted with two different insurance companyservices and told me he's getting eight to 10 inquiries a month andonly about one of them would be a preferred customer.

Obviously, he can't afford to pay more than about $100 a monthfor such a service, when compared to simply buying leads.

Best practices, revisited

It is my opinion that a local agency should view its website asan electronic calling card, a landing page for their social media,and to facilitate communication with their current customer base.They should construct their site to convince those who want tovalidate their professionalism and verify that they're worthy oftrust.

A well-constructed agency website also should be part of anoverall social media strategy, which includes Facebook, Twitter andLinkedIn, that are designed to showcase the agency's strengths.

There are many success stories out there that are legitimate,but they involve large agencies with elaborate websites, operatinga national marketing program.  A website designed for alocal strategy, seeking to reinforce your strengths could be donemuch easier with far less expense.

If you are set on going "all-out" on a national website,Hagerty.com is a greatexample of an online portal that works through a combination ofgreat, extensive content and proper techniques. They provide avaluation tool for classic cars and attract a great deal of trafficunder searches such as "1965 ShelbyCobra." Among other things, Hagerty.com also has thousands ofbacklinks, or links pointed to their site. The average local agentwould be extremely lucky to have several dozen backlinks. Backlinksare considered by most search engine optimization (SEO) experts tobe essential to raising your domain authority.

One can only imagine the amount of resources Hagerty has devotedto its site.  

Is aggregation the answer?

I believe in aggregation. In 1989, I started one of the pioneeragency aggregations. I knocked on over 100 insurance company doors,most of which were politely slammed in my face.

Travelers was the first company to agree to my "new" idea. Ittook a $3 million investment to get our large cluster off theground. In 1992 and 1993 my group was the largest producer in thenation for Metropolitan Property and Casualty.  Thatseemingly opened the eyes of many in the industry to the potentialof aggregation. Now some people suggest that 40% to 50% of smallagencies in the country are part of an aggregation. People arestarting aggregations with extremely small investments because theconcept is now mainstream.

Similarly, I believe the only way agencies can succeed inattracting cost-efficient leads on the Internet is through agencywebsite aggregation. The logical answer would be to bring companiesand agencies together under a mega-site that is independent of anyone company or agency, but implementation of that concept has beenproblematic.

Although we feel our "aggregation" website's content has thecorrect tone and information the user is seeking, I readily admitwe haven't proved the correct formula for this to work. Our agencywebsite aggregation program has been endorsed by the National PIAfor several years. We have received great support from a number ofcompanies that has allowed us to test our theories, includingTravelers, Harleysville. Hartford, State Auto, State Fund Mutual,Mendota and Kemper.

Unfortunately, we've been unable to attract the kind ofinsurance company financial backing needed. More importantly, whenGoogle changed the rules and quality backlinks became necessary tocreate a large amount of traffic, we were not able to immediatelyrespond.

The other agency association website aggregation project,Trustedchoice.com, has been able to attract links from a few dozencompanies and several hundred agencies. They're starting to attracttraffic, but after spending many millions and more than five yearsof work their traffic is a fraction of that enjoyed by Hagerty.com.By some measurement, their traffic is about 5% of what esurance.comis now getting.

So what have we learned?

Website aggregation can work, just as insurance agencyaggregation did. It will work because people want to buy locally,because they know and trust an agent. Small agencies workingtogether in an aggregation attract the large insurance companies todo business with them in an efficient manner. Small local agentscan attract new clients efficiently with aggregated websites.

The traffic is out there. Many large insurance websites' trafficdoubled over the last year. The probability that a local insuranceagency can attract a fair share of that increasing traffic througha cost-efficient program is almost impossible.

Because the returns for SEO are so nebulous, agents need towatch every dollar they spend on their website, after they'vecovered the basics.

Agencies need a website. Agencies need to be active in socialmedia and local search driving people to take a look at theiroperation. They should take part in Facebook, LinkedIn, Google +,and Twitter, for a start. They need a good website to validatethemselves to people checking them online.

Jim Holm, a five-decade insurance industryveteran, is the CEO of Enhancedinsurance.com.