1
00:00:00,790 --> 00:00:03,980
What steps does it take for
someone using your product or

2
00:00:03,980 --> 00:00:06,230
service to get to their goal?

3
00:00:06,230 --> 00:00:10,880
Having these illustrated and tracked
using an analytic service can help you

4
00:00:10,880 --> 00:00:14,990
identify critical breaking points in this
flow and address them strategically.

5
00:00:16,120 --> 00:00:19,910
Let's continue to use
Amazon.com as an example.

6
00:00:19,910 --> 00:00:22,700
First, we need to outline the steps and

7
00:00:22,700 --> 00:00:24,720
the funnel for
making a purchase on Amazon.

8
00:00:26,050 --> 00:00:29,080
Aim for
about 5 to 10 steps in your funnel.

9
00:00:29,080 --> 00:00:30,910
Here's ours.

10
00:00:30,910 --> 00:00:33,674
1, go to Amazon.com.

11
00:00:33,674 --> 00:00:37,245
2, search for a desired product.

12
00:00:37,245 --> 00:00:40,375
3, look at product pages.

13
00:00:40,375 --> 00:00:43,400
4, add product to cart.

14
00:00:43,400 --> 00:00:45,516
5, proceed to checkout.

15
00:00:45,516 --> 00:00:48,538
6, sign in.

16
00:00:48,538 --> 00:00:51,732
7, review check out page.

17
00:00:51,732 --> 00:00:54,120
And 8, place your order.

18
00:00:55,390 --> 00:00:57,710
Now, as you can guess, fewer and

19
00:00:57,710 --> 00:01:01,590
fewer people will go on
from one step to the next.

20
00:01:01,590 --> 00:01:05,230
So really,
it all begins to look like a funnel.

21
00:01:06,370 --> 00:01:11,320
If this visualization becomes hard to
read, you can also use a simple bar chart.

22
00:01:12,550 --> 00:01:15,760
You'll find that one advantage
of using a bar chart

23
00:01:15,760 --> 00:01:19,090
to visualize this flow instead
of the more traditional funnel,

24
00:01:19,090 --> 00:01:23,190
is that it's easier to map the data
while maintaining readability and scale.

25
00:01:24,600 --> 00:01:28,860
One thing that defines all
funnels is they are linear.

26
00:01:28,860 --> 00:01:33,180
In our example where step number 3
is someone looking at product pages,

27
00:01:33,180 --> 00:01:36,750
it doesn't matter how many
product pages someone looked at

28
00:01:36,750 --> 00:01:40,070
since it's all just one step
in getting to the conversion.

29
00:01:41,100 --> 00:01:45,660
The fact that people may look at 3
products or 30 can be significant, but

30
00:01:45,660 --> 00:01:47,580
within a different research context.

31
00:01:48,830 --> 00:01:53,040
Now comes the real useful part
of going through this exercise.

32
00:01:53,040 --> 00:01:58,090
If somewhere in the funnel you
noticed a particularly steep drop-off,

33
00:01:58,090 --> 00:02:00,670
you know it's a problem area
that you need to look into.

34
00:02:01,980 --> 00:02:06,230
As you can see here, a steep drop has
appeared between adding something to

35
00:02:06,230 --> 00:02:09,240
the cart and proceeding to checkout.

36
00:02:09,240 --> 00:02:12,450
People appear to be putting your
product in their shopping carts, and

37
00:02:12,450 --> 00:02:13,450
then leaving the site.

38
00:02:14,560 --> 00:02:17,645
It's quite possible that
this drop-off is normal.

39
00:02:17,645 --> 00:02:22,755
So before you assume something is wrong,
you'll need to gain some context.

40
00:02:22,755 --> 00:02:26,905
Understanding the historical trends for
your product and your industry

41
00:02:26,905 --> 00:02:31,645
will allow you to decide if this drop-off
is worrisome or completely normal.

42
00:02:33,030 --> 00:02:37,610
If in fact you find that
a drop-off is unusually high,

43
00:02:37,610 --> 00:02:42,950
run a qualitative study, such as
a usability test, to understand why.

44
00:02:44,130 --> 00:02:46,740
Once you think you've found
a design solution and

45
00:02:46,740 --> 00:02:48,770
have tested it with a few users.

46
00:02:48,770 --> 00:02:51,630
Consider introducing it
as part of an AB test

47
00:02:51,630 --> 00:02:54,220
to make sure it improves
the conversation rate.

48
00:02:54,220 --> 00:02:57,090
And if the test is successful,
launch to everyone.