A few results of the survey "online purchase habits"

Posted by Anthony Baillard 1 year, 3 months ago

Ordered!In February, we ran a survey about online shopping habits. To thank the 300 participants, we publish some results here.


The survey

The survey was part of a social study based on user-centric design and was online for ten days. It is quite short but was long enough for us to collect 290 answers from 35 countries all around the world. The study aimed to understand the habits of users in order to provide them the best experience.

We'd like to share a few graphs among the ones we plotted in order to highlight some interesting discoveries regarding online shopping habits in this trouble times.

Let's start with an estimate of the representativeness of our sample. We compared the age distribution of the participants to the age distribution of Internet users worldwide (statista.com).
The grouped the data by generations for meaningness and got a good surprise: the distributions are really close (figure 1)! It means we can be quite confident about the representativeness of the results even if 300 answers are far from enough to make a worldwide generalization. The following plots cannot be used for a rigourous analysis but are interesting guidelines for further studies.

A reminder for the generations:

  • Boomers (1946-1964), from 57 to 75 years old
  • X (1965-1980) from 41 to 56 years old
  • Millenials (1981-1996) aged 25 to 40
  • Zoomers (1997-2012) aged 18 to 25 (children under 18 are not counted)

Distribution of generation (survey)Distribution of generation (worldwide)
Figure 1: on the left, the age distribution by generation of the 290 participants of the survey. On the right, the age distribution by generation of Internet users worldwide (source: statista.com). Zoomers are slighly underrepresented in the survey.


Online shopping habits

Here is the meaning of the frequencies to understand the plot of figure 2:

  • Everyday (Once or several times a day)
  • Often (several times a week)
  • Regularly (Every week)
  • Rarely (Once or twice a month)
  • Almost never (Once or twice a year)

Most people buy between once a week and once a month (~75%). It is an interesting result, leading to the idea that buying online, even during lockdown is not as frequent as one could think. These numbers are interesting to keep in mind to analyse figure 5 (see below). Figure 3 suggests there are two different ways of buying online: using a computer or using a smartphone. The very poor results of tablets could mean that smartphones are used in specific contexts, when people are outside or on the move but we have no proof of that yet.

Purchase frequency
Figure 2: Purchase frequency. 75% of the participants buy Rarely or Regularly, that is once a week or a few times a month.

Figure 3: Preferred device when buying online. There is no significant different between genders, the preference being equitably shared between Desktop (computer) and Mobile (smartphones).


Figure 4 gives an interesting insight of the usual user journey between the moment of the desire to buy and the validation of the bought. Search engines are the main entry points, conforting their importance in nowadays consumption societies.On the other hand, more than 30% of the participants look for a product following a social media or friend advice, suggesting that social relations are also an important driver to buying decision.

Nevertheless, most people end up buying on marketplaces or website of known brands, which are generally huge platforms. It also suggests that most people are careful when buying online, avoiding untrusted websites or small sellers.

Figure 4: On the left, the distribution of searching strategies, that is how the participants usually start looking for what they are going to buy online. Search engine is by far the most common answer but it is interesting to note that friends advice (that is not necessarily digital) reaches a good score. On the right, the distribution of purchasing platform. Marketplaces and Brand websites you know gather 80% of the answers, suggesting huge platforms are preferred.


Commitment and pandemic

The last serie of plots (figure 5) is really interesting and would deserve a more detailed study. Our analysis of these four plots is that people might feel more commited that they really are. Indeed, around 2/3rd of the participants would say that the pandemic has had an influence on their online purchase habits but only 1/2 of them have already bought something helpful for others. 

Figure 5 (from top to bottom): first plot, the effect of the pandemic on the participants' habits. 7 people out of 10 say that the pandemic and its consequencies have influenced their online purchase habits. Second plot, the ratio of participants that have at least once bought something to help someone in need is around 50%. Third plot, the ratio of participants willing to buy something slightly more expensive to support a cause is around 75%. Last plot, about 60% of the participants affirm that the pandemic has strenghtened they personal commitments.





Learn more

Link: pandas, the python library used to plot the graphs