Large Differences Between Top and Bottom
The general average for page weight among the hundred sites investigated lands at 9 MB. If Icaniwill – which is considerably heavier than the other sites in the comparison – is excluded from the calculation, the average is just over 7 MB, while the median for the entire survey is around 5 MB. The differences between the extremes are extensive; the heaviest site in the survey is 131 times heavier than the lightest, Svenskt kosttillskott. Icaniwill also weighs more than the 52 lightest sites on the list combined.
Industry Patterns are Characterized by Product Range
The analysis shows that different industries have fundamentally different approaches to how much data is transferred, which can often be attributed to how the products are marketed:
- Fashion and Lifestyle: This category is generally the heaviest in the survey, with an average of 12.4 MB and a median of 7.8 MB. This is almost double that of the electronics industry. The result reflects the fact that fashion is largely sold using high-resolution images and visual experiences.
- Electronics and Technology: Players in this segment are distinguished by delivering lightweight websites. Dustin, Netonnet, Inet and Komplett all lie below 4 MB in transferred data. Here, the focus is primarily on specifications and technical data rather than aesthetics, which is visible in the page weight.
- Pharmacy and Health: Within this segment, the spread is large. Apotea loads 2.8 MB, making it one of the most efficient sites in relation to its size. Apoteket Hjärtat is, in contrast, twice as heavy with 7.7 MB, despite a lower turnover. Lyko and Meds are positioned in the middle.
- Nicotine Products: Snus sites such as Snusbolaget, Snusme and Nordicpouch are generally light. This is partly explained by the fact that the products are visually uniform and that strict regulations regarding marketing limit the use of extravagant design elements.
Efficiency Metric: Turnover Per Megabyte
When page weight is related to companies' turnover, an efficiency metric emerges where the median for turnover per MB is 131 million SEK.
The most efficient sites are characterized by high turnover in relation to page weight. This category is dominated by large companies with well-optimized and conversion-focused platforms, including Apotea, H&M, Netonnet and Dustin. Amazon is also among the top ten in this regard.
At the bottom of the survey are websites where aesthetic ambitions and brand building are allowed to cost in the form of increased data volume. Brands such as Icaniwill, Strongerlabel and Aim'n are located here.
Top 100 E-Commerce Companies Ranked by Weight
Place |
Domain |
Turnover (MSEK) |
Average Transferred MB |
1 |
329 |
184.33 |
|
2 |
1727 |
41.73 |
|
3 |
385 |
36.53 |
|
4 |
3400 |
36.17 |
|
5 |
297 |
30.07 |
|
6 |
404 |
26.70 |
|
7 |
1373 |
24.90 |
|
8 |
39455 |
18.20 |
|
9 |
1352 |
17.20 |
|
10 |
8037 |
16.43 |
|
11 |
440 |
15.37 |
|
12 |
249 |
13.40 |
|
13 |
443 |
12.50 |
|
14 |
1094 |
10.80 |
|
15 |
304 |
10.40 |
|
16 |
695 |
8.50 |
|
17 |
336 |
8.47 |
|
18 |
3592 |
8.30 |
|
19 |
1006 |
8.30 |
|
20 |
1425 |
8.20 |
|
21 |
695 |
8.13 |
|
22 |
371 |
8.00 |
|
23 |
517 |
7.90 |
|
24 |
416 |
7.90 |
|
25 |
411 |
7.80 |
|
26 |
2511 |
7.70 |
|
27 |
829 |
7.40 |
|
28 |
1900 |
6.83 |
|
29 |
291 |
6.83 |
|
30 |
305 |
6.80 |
|
31 |
648 |
6.77 |
|
32 |
12244 |
6.70 |
|
33 |
864 |
6.57 |
|
34 |
540 |
6.53 |
|
35 |
415 |
6.30 |
|
36 |
1910 |
6.20 |
|
37 |
292 |
6.10 |
|
38 |
1302 |
6.10 |
|
39 |
406 |
6.00 |
|
40 |
373 |
6.00 |
|
41 |
2201 |
5.87 |
|
42 |
323 |
5.80 |
|
43 |
357 |
5.70 |
|
44 |
524 |
5.67 |
|
45 |
2612 |
5.53 |
|
46 |
634 |
5.50 |
|
47 |
1633 |
5.47 |
|
48 |
432 |
5.23 |
|
49 |
278 |
5.23 |
|
50 |
399 |
5.10 |
|
51 |
1689 |
5.00 |
|
52 |
308 |
5.00 |
|
53 |
1655 |
4.97 |
|
54 |
238 |
4.70 |
|
55 |
2020 |
4.63 |
|
56 |
272 |
4.60 |
|
57 |
1145 |
4.57 |
|
58 |
574 |
4.40 |
|
59 |
281 |
4.40 |
|
60 |
684 |
4.33 |
|
61 |
795 |
4.30 |
|
62 |
3600 |
4.20 |
|
63 |
1375 |
4.10 |
|
64 |
320 |
4.10 |
|
65 |
718 |
4.07 |
|
66 |
973 |
3.93 |
|
67 |
893 |
3.90 |
|
68 |
409 |
3.90 |
|
69 |
474 |
3.77 |
|
70 |
351 |
3.70 |
|
71 |
606 |
3.60 |
|
72 |
1229 |
3.57 |
|
73 |
798 |
3.50 |
|
74 |
965 |
3.43 |
|
75 |
2700 |
3.40 |
|
76 |
707 |
3.40 |
|
77 |
990 |
3.20 |
|
78 |
813 |
3.10 |
|
79 |
414 |
3.10 |
|
80 |
1589 |
2.97 |
|
81 |
5872 |
2.90 |
|
82 |
713 |
2.90 |
|
83 |
663 |
2.90 |
|
84 |
386 |
2.90 |
|
85 |
2724 |
2.87 |
|
86 |
846 |
2.87 |
|
87 |
1949 |
2.83 |
|
88 |
6408 |
2.80 |
|
89 |
1817 |
2.80 |
|
90 |
1398 |
2.70 |
|
91 |
286 |
2.70 |
|
92 |
260 |
2.70 |
|
93 |
1506 |
2.63 |
|
94 |
297 |
2.63 |
|
95 |
758 |
2.60 |
|
96 |
255 |
2.33 |
|
97 |
824 |
1.90 |
|
98 |
348 |
1.80 |
|
99 |
799 |
1.60 |
|
100 |
370 |
1.40 |
How We Conducted the Survey
To map page weight, all one hundred websites were tested via the Chrome browser in incognito mode with the "Disable cache" function enabled in the developer tools. On the first visit, all cookies were accepted and any pop-up windows, such as country selectors or currency settings, were closed.
Thereafter, three reloads of each site were performed. The values from the very first load were excluded from the final result. The reason for this is that some websites show extremely low and potentially misleading values on the first visit.
The average of the three subsequent reloads formed the basis for measuring "Transferred" (transferred data) and "Resources" (total amount of resources). Data was read only when the page had completely stopped sending information. In cases where moving elements, such as image carousels, loaded new data with each image change, the measurement was waited for until all images had been downloaded.
During the tests, it was noted that the clothing brand Icaniwill showed deviating results and the measurement therefore had to be repeated more than three times. It turned out that a single video with a product on the homepage loaded up to 20 MB with each restart. At most, the page transferred 1.5 GB of data before the test was aborted.