Home

Dating apps and the secrets of machine learning?

It’s not big news that I’ve been occasionally single, traveled around or lived in other countries and used different kind of dating apps. Recently I read an article ”The Tinder algorithm, explained” and another one about ELO scoring system they are said to be using. It came obvious for me that there actually not exist articles about these companies and their use cases for machine learning for better value for themselves. People who use dating apps are more likely to click digital advertisements according to my colleague and finding these people is one of the best ways to get higher click rates and ROI.

 

It’s a beautiful thought that Tinder, MeetMe, Skout or any other application concentrating in online dating would exist to make perfect matches to help people to find love of their life. Maybe over 85% of men who use dating apps are into casual relationships and I believe female wouldn’t be much different. And about the companies behind these apps, do they get paid for the rate of long lasting relationships, marriage with kids or other KPIs which would make them think how the app can better help in the beautiful thought? Absolutely not. Actually it’s opposite, each relationship would loose one existing user and to be honest, most people are not ready to pay for usage and then it goes to advertising inside the app.

 

Basically I claim that machine learning is used for better targeting valuetion of each and every user. It can be either to buy a premium membership (not sure about perentage who pay for usage, but it’s easier money in the business case) or better advertiser value. These are the cases I take into closer look.

 

What kind of users would buy a membership? People who are really interested in finding their soulmate or singles living a playboy life, then it’s OK to hide advertisements from them (already cashed in). But for the advertising it’s said that Facebook is actually trying to irritate people, some trolling or such will make users more likely to click an advertisement. For dating apps I would consider this differently: a) men who are good target for advertising are not satisfied about the situation, are willing to change their life or buy a pill which makes them more desirable for female and b) female who’s falling in love is more likely to want to keep the man and uses more money to buy anything to get a ring after successful dating.

 

Now the big thing: machine learning needs to be used to help in targeting advertisements. For example Tinder it’s useful to keep people swiping, does it? Certain amount of profiles, then an ad and swiping continues, new ad and more swiping. Keep this going. Men and female can be disappointed in different ways: a) no matches to men, they keep swiping (otherwise they would use more time to get in bed with the new match) and b) for female it’s better that they disappoint after the match to keep them swiping (we must be honest, new male matches will ask for date, come their place or something very soon and if the discussion goes casual things match will be removed). So irritation is easy to make. It’s not rocket science.

 

Okay, there still need to be matches for everyone and these apps need to think which targeting is useful and which is not. I believe because what I told about men and female buying behavior, it’s OK for female to get match in their home location (they will shop more) and for men it’s okay to match outside their home location because these locations selling the ad for someone who doesn’t live in the location is not so valuable.

 

I believe from personal experience these make sense. I usually get a lot of matches while I’m traveling, but in hometown or places I have lived before not so likely (usually you give access to your Facebook and then this information is available). Most matches in cities I lived were from someone who’s just visiting relatives nearby but not actually live in the city. For sure some markets in general generate more value for potential advertising sells. For example South East Asia the average income is lower than in US or Europe or North Asia, also India is quite new to digital marketing and 95% of population are still living around 11.000 USD annual income. This will have some impact in algorithms and business models as wells.

 

The countries you use the app have always different behaviors. For example it’s more casual in Europe to tell going to new date with new person many times a week while in many Asian countries culture encourages to get in long-term relationship. Naturally this will affect how many people and what kind of people use the app in the country. Adventurous, interested in meeting new people is a must, but if your country’s culture is more family-oriented you might like to meet foreigners much more than locals from apps. And if the local culture is against mixed kids, it’s more likely your values and potential dates come from local origins. Dating apps need to localize their products for their algorithms as well. I believe this is another aspect for them to make more profits.

 

As a conclusion my claims are

1. Machine learning is not used so much for potential matches, but value from customers (premium membership or more clicks to ads)

2. Location outside your neighborhood you’ll be less interesting for advertisers

3. Men and female are different in buying behavior, aka advertising click rate

4. It’s not useful to make perfect matches, people will stop using the app in long-term relationship

5. We’re still using the most of the apps for fun, don’t take it personally if you are not feeling to be so popular or getting enough matches

6. Machine learning to offer better quality for app users and advertisers is not an easy thing, people need to be kept partially happy to keep using the app, but partially irritated to make clicks for advertisers.

 

From my personal experience your look shouldn’t matter so much in any location you try to find a date. Basically our thinking comes from local culture, habits and after all entertainment industry, movies we watch or see in magazines. It doesn’t then make any sense that in some countries you get almost all likes a match and other country where you swiped doesn’t have any impact. There must be also the classification also who sees your profile, who would you likely to match, but not for the reason to get more matches, the reason is how to avoid it. I suggest to try fake your hometown location, locate your phone mostly in other city and occationally check your hometown profiles. This way in my theory you should be able to get more dates nearby and potentially meet someone. If there ever will be questions a) did you two date, b) was it succesful, maybe better not to answer or speak true. It would be used to check algorithm is working, not the happiness of user.

How to success in your daily decisions?

I’ve been studying and thinking about efficient decisions making recently. To put it simple decision making is just a process where we decide between options and value each option to find the best outcome. For example Psychology today mentions how important our internal emotions, feelings or instincts can affect to the wise outcome.

 

There’s also discussion about decision fatigue. Each day we make approximately 35.000 decisions. As in this article, I have adapted certain habits which help my daily or weekly routines: 1. I usually cook more once a week, like lunch boxes, breakfast or snacks I take to office, 2. I established a routine for workout and sleeping time and 3. I try to avoid too much time in simple decisions.

 

In the era of AI, machine learning and other trending topics, humans are no longer needed in simple decision. It’s surprisingly still how much we put on supervisors or line managers to decide very simple things like holiday approvals or travel plans. If there are rules, processes and other guidelines in place we should trust more on each individual to follow them. When companies still spend time and effort to control all these activities, it’s easy to notice decision count will never go lower.

 

What’s the best time to make decisions? The article I mentioned says morning is the most efficient time. If you take into look business negotiations or politics, why do meetings still last very long? Maybe is strategy or tactics from one side to keep it going, avoid more thoughtful decisions when all people are on their best. You can also look at how salesmen work when you consider buying something new for example cell phone. They try to tell more and more details that will make you more tired, maybe easier to go their direction. Keep calm, have reasonable options before your decisions and it will certainly go for more meaningful direction and result good decisions.

How to manage business negotiations successfully?

I read recently two interesting articles. One was about cross-cultural business negotiations and another about managers who are best suitable for cross-cultural positions or work abroad.

 

Business negotiations which involve people from different backgrounds, cultures and nations we often face a fact that we have our own expectations, stereotypes and ways to think before even meeting a person. An interesting finding was that actually best negotiators can confront these facts and be more successful in getting a result. Counterpart also must concentrate more when they not just communicating with similar backgrounds, same national counterparts.

 

You need to be able to meet people as individuals. Make your research to be successful negotiator. This happens easiest by giving some time to learn co-operation, learn about expectations. Exactly similar preparations as you do for all meetings or negotiations. You should also know about the culture, ways to do business and find solution. This can happen by time but is always doable.

 

Also negotiations may be tough for each participant. Since you already get to know in early phase, all people involved are actually better prepared to meet face to face. In project environment I prefer to organization real kick off meeting, even it’s not always necessary to go through pure facts, schedules and such. It’s important process to learn co-operation.

 

Key characteristics for successful cross-cultural or expats included attitude to change (it’s normal, positive and source for opportunity); ability to think creative, outside-of-the box way;  know-how how to change also yourself; expertise in emotional ability for transition (maybe also emotional knowledge); and great learning skills for new ways of thinking. I can recognize myself in those; I have worked before with different nationalities and lived in other countries for longer time. It always helps to in adapting new living environment by meeting, getting to know locals and local ways to eat go out, and so on.

 

Cosmopolitan minds are natural travelers; they are interested in moving around and getting to know. This is a skill you can learn. Let yourself explore, put yourself out of comfort zone and realize what the world has to give you.

Successful multicultural teams and people

I read recently an article about successful multicultural teams. It actually mentioned that this type of teams can be almost 30% more efficient compared to other teams. For companies it means they should have more diverse teams, more diverse people and try to find a good mixture. You can read full article here.

It's not a coincidence what sort of people are ready to move in different places and join new teams. It's more than likely that people who do it are sure about their emotions (know themselves), open to new things (easy to get adapted) and in general social (extrovert). If you have this kind of people in your organization and need new experience to overseas market, consider people's willingness to relocate and then about knowing the person who is planned to move, join and integrate. It's a good match to be made.

Another thing to pay attention is knowing that people who relocate or join multicultural teams are known about their easyness to get along with different people but also to know what's socially acceptable in different cultures. They need to be prepared not only for people but local customs, ways of living and long history, cultural learning. Usually when you travel you're able to gain this kind of experience, openness and easyness and maybe make some friends already. Before I moved to Taiwan, I actually got to know the place and got to know the people before my arrival. This helped me a lot to easily adapt in my new country, neighborhood and job.

If you're interested more about cross-cultural or multicultural management, pay attention to my other articles in this website.