Thinking of joining eHarmony? Clearly eHarmony sells itself as the best website for matching singles. They claim they have millions of users, use a scientific treatment to match singles, analyze 29 dimensions of compatibility, and have hundreds of success stories. Good, science I understand well, I’m just having trouble watching the science through all the business. Like any good scientist, I question the facts as introduced to confirm or invalidate their findings. Not only are my results are very different than advertised, I believe I have studied the behavior of the eHarmony system enough to understand some of its internal workings. What I believe I have found is very disturbing.
If you are fresh to eHarmony, it will very likely help if you understand some of their terminology. They have a page that describes the communication process, another that lists the 29 Dimensions of Compatibility, and even have a page that talks to their Scientific Matching. If you are more nosey, you can take their Guided Tour, which I consider to be unspoiled marketing hype, and will explain shortly.
eHarmony is unlike any other site out there. When you very first join, you are required to finish their 436-question Relationship Questionnaire. How swift do you think you can reaction all of these? Recall, you need to wait for the web page to geyser, read the question, give some thought to your response, scroll the mouse to the correct button, and click. I would estimate this to be toughly 30 seconds/question or 1/Two a minute. 436*1/Two=218 minutes (over Three.Five hours). Most places would provide you with some estimate of time before you commence – that was certainly missing when I joined. After Four hours of answering question after question, I was eventually able to log in and begin searching for matches.
October 22, 2005 I pressed search for matches in the age range of 22 to 32 years of age (as recommended by eHarmony) within a 60-mile radius of my zip code. No matches found. A day or two later, I received five emails of the following format.
There’s someone we’d like you to meet.
Dear Barry and So-and-so,
eHarmony’s Compatibility Matching SystemTM has found that you two are a very compatible match, based on our rigorous 29 Dimensions scale. With this level of compatibility, we believe you two will have a lot to talk about and may even find true love together.
So, Barry and So-and-so, we’d like to make an introduction.
- So-and-so lives in Hamden, Connecticut United States. Read more.
- Barry lives in Stonington, Connecticut United States.
The next step, Barry, is for you to begin communicating with So-and-so. Explore your collective areas of compatibility, and have joy discovering if there’s chemistry.
Recall, your match has been cautiously selected from millions of candidates because you share extreme levels of compatibility in the areas proven to create relationship success. So we encourage you to consider this match with care and find out what lies beyond your summary profiles.
The next step is simple—just begin communicating today!
Dr. Neil Clark Warren
Of course, this is an interesting letter. It implies that she received a letter telling her that the next step is to wait for me to contact her, which is not true. However, being matched up with someone means that you may both see the other’s profile. I reviewed my matches’ profiles and determined that I should add some more detail to my profile. I was leaving for business travel for several weeks so I requested that eHarmony not find any fresh matches for me while I was away so I could add more content. While on travel, I received the following message:
Congratulations! eHarmony has found a very compatible match for you and they have already sent you their very first communication.
So-and-so from Hamden, Connecticut, United States has reviewed your basic information and would like to embark the process of getting to know you better. You can find out more about So-and-so and response some elementary questions by returning to www.eHarmony.com. Click here http://www.eharmony.com/singles/servlet/login.
We encourage you communicate with So-and-so using our effortless, safe and anonymous Guided Communication process. Let eHarmony guide you as you get to know your match from the ",inside-out.",
For more information about communicating on eHarmony.com, please visit our Help page.
The eHarmony Team
I clicked on the user and it said I needed to fully join to begin communication. Before you get to this point, the only place I have seen pricing information is from the last tour page, which only tells you that being part of the community is free, and that the personality profile is a $40 value. I’ll let you in on a little secret. The service charges $50/month, which is toughly twice what every other dating service charges. I was coaxed by the marketing type that this was the best place to meet singles, so I determined to join for three months (seems like a fair amount of time to test it out).
I answered her questions, and have not heard back since. Perhaps she is not interested based on my answers, or perhaps she is no longer a member. I have no way of knowing. Like all dating services, once you create your profile, you are part of the community (hey, it’s free). This doesn’t mean you are a paying member and can react to anyone, it just means that people can see you. Most services are nice and let you know that the person you are attempting to communicate with is not presently a member so you are not too astonished when you don’t hear back from them. The good services also let you know if someone has not logged in for a long time indicating that there is a strong probability that they will also not react. No harm in attempting, just don’t be too hopeful or disappointed. With eHarmony, they do not tell you anything except that they think you are a good match. I have determined to arbitrarily draw the line at 30 days (and this includes the time before nudging matches which I determined to do at harshly 20 days). This has been the case with toughly 33% of those I have been matched up with. Please note, I computed this value by going back 30 days and comparing the number open at that time relative to the number of people who I now know will not react for the next 30 days. This should permit me to predict how many of my current matches I will never hear from – I just don’t know which ones.
Another common occurrence is I get an email from eHarmony indicating that they found a fresh match for me. I check my email daily and log in upon receiving a fresh email. However, I have frequently logged in to find that she closed the match already. This occurs harshly 20% of the time. Perhaps my Personality Profile is indeed bad or something. I had several friends review it and they thought it sounded relatively accurate.
Matches get to provide a reason why they are closing the match. The most popular reaction I get is ",distance inbetween us is too good.", Notice that physical distance and likelihood of moving for a match are not part of the 29 Dimensions of Compatibility. Based on my practice, this very likely should be since for many of us, at our age and current status in our careers, we are unlikely to pick up and budge for someone. I’m not sure eHarmony fully understands this. In my profile, I desired to have them restrict my matches to within 30-miles, but I realized that I live in an area with very few singles and wished to expand my radius to pick up the near-by cities (Hartford and Providence). I enlargened the radius to 60-miles which is the next larger choice. To me, this means I hop in my car, obey the laws of physics, and 60-miles later, I say hello to my match. Please note, I do not own an amphibious car. Hopefully they are selecting closer matches for me very first, but I already mentioned that I do not know how many people they have in my area. Determining distance to a match is fairly straight forward using any mapping service such as Google, yet eHarmony uses the more classic style of ",as the crow flies", which results in matches that are closer to 180-miles away, rather than 60-miles. It is a petite wonder why I get rejected on distance as often as I do. The following map shows where all of my matches have been. The inward circle is a 30-mile radius. The shaded area represents 30- to 60-miles. I must admit, I’m astonished by the number of matches outside of the 60-mile radius.
The most interesting closed match is the one that eHarmony closed by accident. As usual, I received an email indicating they found a fresh match, logged in, and sent the very first questions to her. A few days later, I noticed that she was not in my list of people I was communicating with. I went to my closed page and spotted her listed. I clicked to see what the reason was for the match being closed and found that I had closed the match and that the reason was an invalid string. I contacted technical support who told me that I had closed the match (I most certainly did not), a glitch had occurred in the database so no one knows why the match was closed, and that they would attempt to contact her to see if she was still interested in communicating. Weeks have gone by, and I’m certain I will not hear from her again. It wasn’t until I discovered extra bugs with eHarmony that they admitted I did not close the match and that they have a known bug they are working to resolve.
The 2nd problem I had with eHarmony was attempting to reply to a match (my very first open match) that wrote to me through their on-line email. I did exactly what they recommend. I wrote my response in Word so it would be spell checked, copied the text into the reply text box, and clicked send. Of course, it didn’t actually send. I tested the behavior to ensure it was repeatable and it was. I have a master’s in computer science, and have worked in tech support before. I wrote an email to them summarizing the problem and included information about my computer, operating system, etc. The response I got was a standard form asking most of the questions I already answered in my very first email. I responded by packing out the form and mentioned that this was included in my very first email. We went back and forward with them asking me to turn off firewalls (which I didn’t have on) disable virus scanners (which I don’t use) enable cookies (which already were). At no point did I get the impression that they actually attempted to create the problem using the information I sent them. They did send messages to my Two matches from my account which looked like they came from me. Had they followed my instructions they would have recreated the problem, but they did not listen. So messages went to my matches telling ‘test do not reply’ and nothing else. I had asked them to set up a fresh test match so they would not interfere with my attempts to communicate with my real matches. They said it was not possible to do this. I ultimately asked them to call me since it was over three weeks and we had not solved the problem. Their reply said they would call me within Two days. I heard nothing for Five days and determined to remind them that they owed me a phone call still. Ultimately, a week later I got a call from them. Very first question – what operating system do you use? It was like they had never heard from me before. To make it even more annoying, I was talking with the same individual I had been emailing! After almost an hour on the phone, we confirmed that I had found a bug. So for a month, I had been incapable to reply to matches using their recommended treatment. Fortunately, I was able to figure out some work-arounds, and after explaining the random messages they received, all hope was not lost.
For being so patient on both issues they determined to extend my membership by 14 days. This prompted me to write this web page. Based on how they treated this issue, I have to wonder if they sent a message to the match they closed. If they did, did they take responsibility for their mistake or did they simply tell her that I dreamed to reconsider.
With the remaining set of potential matches, I get the chance to attempt to make my way through the guided communication to open communication. To date, only Trio women have bothered to pursue communication. In each case, it took approximately Two weeks to go from being matched to being able to email. Do I feel that these are good matches? Not at all. What do we have in common? We both like to listen to music, love live music, and value friendships. Amazing, I didn’t think anyone else out there had these interests. Clearly a strong foundation to build a lasting relationship. As we have talked, there hasn’t been too much else we’ve found in common. All have been very nice, polite, and conversational. If they lived a little closer and we had more in common, we might even become good friends.
Humorously enough, using myspace, I’m meeting more people who I share more interests with and are within 30-miles. I know we joke that Fresh Englander’s measure distance as time, but 30-miles is generally 50-60 minutes of driving time. 60-miles is closer to 90-120 minutes depending on highway access for the journey. In fairness, eHarmony does suggest something they call FastTrack to speed up the process, but given the high number of closed matches and no-communications, I don’t want put pressure on anyone. Seems reasonable thinking to me.
Alright, back to science. Let’s look at some of this information graphically to get a better feel for how it is all coming together. Looking over time you can see the cumulative matches created (blue) compared with the cumulative number matches that are not closed (pink). Of course, I consider matches that are over a month old with no communication to be inactive. Therefore the number of active matches (yellow) are the women I might actually be able to communicate with.
A few observations about this graph. Very first, notice that the blue line is almost linear with a match rate of less than two matches created per day. I know the excuse they give is that they are searching 29 dimensions in an effort to find that special someone and that takes time. No it doesn’t! Database queries (even on poorly configured systems) do not take this long. They tout having over 8 million users, which tells us the size of their index table. If you want to search 29 dimensions, you very first search one dimension. This reduces the number of records you have to filter out with the next dimension, which makes the search go swifter. Even if you have to search every record 29 times, this still doesn’t take more than a few minutes. Don’t believe me, head over to Google. Granted, Google is able to process queries quicker than anyone else in the world. So let’s be fair and assume that eHarmony is 1000 times slower. I believe the query should still be almost as rapid as Google if not swifter. Very first, eHarmony only has 8 million records to search through compared with the over Four billion records Google searches. Even more importantly, eHarmony knows they will be searching by geography. They should have cracked the country into regions and pre-filtered people into those regions. To search, they only have to query regions that fall within the requested radius. We’re not talking rocket science here. I would consider this to be a reasonable semester project for a single junior or senior in college to finish. Again, I digress. The pink curve shows that of these matches, less than half remain open reducing the number of matches per day or less than one. Notice this curve has reached a sustained state. Matches being created and closed are happening at almost the same rate. The most interesting observation is the yellow curve that shows the number of active matches has already stopped enhancing and is decreasing. This implies that matches are closing quicker than fresh ones are being created. This generally shows the end of an experiment. I would tend to conclude that I reached the peek potential for a match around mid-to-late January and I’m reaching the boundaries of the number of users they have signed up.
What if we look at the data a little differently. We know the driving distance to each match and we know the date of the match.
Notice that early on, eHarmony attempted to find closer matches. It looks like in early January they had finished matching me up with all of the users they had in my local area. From that point forward, it looks like the distance to matches has been steadily enlargening so they can keep finding fresh potential matches. But recall, the majority of matches are closing me for reasons of distance. The enhancing distance trend combined with the dwindling set of active matches adds support to the conclusion that the eHarmony experiment is coming to an end. Now this begs the question of what will they do when I totally run through all of the users in the system? Will they just embark matching me with anyone even if they have nothing in common with me? Will they tell me that they have no more users and refund the remainder of my membership?
I called customer support find out how they treat memberships. They only ensure 12 matches for a 1-year membership. They didn’t say how compatible they need to be, just that they would meet your age (minimium of 6 year range) and ethnicity criteria. Humorously, they were not able to tell me how many people are in my area, but they did confirm that the system does not factor in water when determining distance. Another interesting fact they mentioned is that only a portion of their database is searched nightly. This means that you most likely won’t get the ",best", matches very first. You get the ",best", randomly searched subset.
Still think that eHarmony is looking out for your best interests? Consider their compatibility criteria and factor in some of their marketing hype. Overlook the details and assume a uniform random distribution for people joining eHarmony. Assume the same distribution for inhabitance across the country (clearly incorrect, but for purposes of estimating valid enough to give us a good ballpark response). Note: I’m disregarding the fact that eHarmony has international users – I’m including them as in the USA. At present, they claim over 8 million users, so let’s assume 9 million. The united states consists of toughly Three.Five million square miles which puts the eHarmony density at toughly Trio people per square mile. I have elected for a 60-mile search radius which covers about 11,000 square miles. Of course, living near the coast a good portion of the radius is in the ocean. In theory eHarmony will have about Five,000 users in the area. Next we have ages, and I’m willing to assume that users are evenly distributed from Legitimate to 60 years of age. This means my age allowance covers one-third of the users leaving us with a little better than 1,500. We do not know how many choices exist for each of the 29 dimensions, but let us assume the best case screenplay (for eHarmony, not for me) that each dimension has Two choices (i.e. yes/no, not crimson/green/blue). This means that there are 536870912 possible ways to select answers for each dimension. Let’s assume that we only care about 9 of these dimensions (less than a third) then there are still 512 ways to response. Assuming uniform distribution again, this means that eHarmony might have Three people that meet a few of the dimensions that will be compatible with me. Not very good odds. After that, they are sending me anyone they have within my selected radius (or just outside it when they cannot find anyone within the selected radius). I could make the assumptions above more realistic, but the odds that eHarmony actually has a potential match gets lower and lower. For example, I do not live near a major city and therefore have a significantly lower population density. I also think matches should very likely have more than 9 (of 29) dimensions that they are compatible. Like I said, the odds are not looking very good. As a sanity check, I recommend looking at OKCupid which also uses criteria to create matches, is a free service, and provides excellent information and support. By the way, they presently have a total of 1,284 females inbetween the ages of 22-36 within a 50-mile radius of Stonington, CT.
Like any scientific research, I expect people to do their own exploration to see if my results are the norm or an anomoly. My findings suggest that eHarmony does an excellent job of manipulating what information you are provided in an effort to maximize their business – not your happiness. Just in case you get matched up with me on eHarmony – don’t expect to hear back from me. When my membership runs out, I will undoubtedly not be renewing. Look for me elsewhere.
Either way, I wish you the best of luck in your searches,
ps In case you are wondering what I’m looking for, here are the basics.