272709_en-US

liliconcom

Content Archiving

As part of our "information findiing" initiative. we need to look at the current content within the Community. In 20.3, Khoros introduced Content Archiving. Some notes regarding items to consider and potential steps to take within the JMP User Community.

 

From a Khoros product manager on the About Content Archiving article:

 

Ideally, the content which should be archived must satisfy at least one or more criteria from below:

1. The content is obsolete - content having old information, which is not of much use to members anymore
2. The content has a more recent, fresh version available - another topic/article which contains same content but with more recent information
3. The content is inactive - rarely anyone visits (views) or performs any activity (kudo, reply etc) on this content

Even with the above criteria, the decision to archive should be taken after considering the following:

1. Does the page still generate sufficient organic SE traffic?
2. Does the page still generate sufficient 3rd party site traffic?
3. Does the page still generate, or participate in, any significant leads or conversions?

NotesRe: Content Archiving

As of 16JUN2020,

 

The data for Discussions only includes topics and not the reply to the topics. Only a topic can be archived. When a topic archived, the replies are also archived.

 

Category	# of Messages
blogs	1985
discussions	12106
fileexchange	469
insidejmp	2967

 

Thoughts on criteria used to determine if content should be archived:

• Number of views
• Date
• Kudos
• Other?

If we look at views alone:

 

Category	Mean	Quantiles5	Quantiles10	Min	Max
blogs	2530.1964736	514.3	801.4	13	51228
discussions	4221.0498926	347	645	11	76966
fileexchange	5037.4840085	937	1770	66	51380
insidejmp	36.306370071	1	3	0	6303

 

This does not account for older messages most likely having more views than more recent messages.

 

So, let's account for date. Technical Support will officially only support the current version and the version prior. At present, let's assume that posts since 2018 (JMP 14 release year) are more likely to be relevant to supported releases.

 

 

	Category
Year	blogs	discussions	fileexchange	insidejmp
2007	56	0	0	0
2008	93	0	0	0
2009	106	299	0	0
2010	106	480	0	0
2011	119	401	0	0
2012	163	543	0	0
2013	158	545	0	0
2014	117	651	180	256
2015	274	1140	96	381
2016	177	1230	58	378
2017	133	1605	47	365
2018	148	2162	39	627
2019	199	1893	31	446
2020	136	1157	18	514

 

	Category
supported_release_year (2018 or after)	blogs	discussions	fileexchange	insidejmp
no	1502	6894	381	1380
yes	483	5212	88	1587

 

Obviously there is content prior to 2018 that would still be relevant. Views is a good indication. Another potential data point would be the number of kudos received.

 

	kudos
category	Sum	Mean	Std Dev	Median	Min	Max
blogs	3665	1.8463476071	3.5023131771	0	0	30
fileexchange	958	2.0426439232	2.7716309808	1	0	27
insidejmp	1338	0.4509605662	1.3418687921	0	0	27

*discussions were omitted because the data only included topics

 

In looking at a combination of views broken down by year for each category

 

	category
	blogs					discussions					fileexchange					insidejmp
	metrics.views					metrics.views					metrics.views					metrics.views
year	Mean	Median	Quantiles10	Min	Max	Mean	Median	Quantiles10	Min	Max	Mean	Median	Quantiles10	Min	Max	Mean	Median	Quantiles10	Min	Max
2007	1145.9464286	1020.5	721.1	576	3453	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.
2008	1319	993	758.2	550	16280	.	.	.	.	.	.	.	.	.	.	.	.	.	.	.
2009	1304.4245283	1023.5	760.1	656	7044	3430.0234114	2151	1702	1525	17723	.	.	.	.	.	.	.	.	.	.
2010	2385.245283	1389	887.7	609	51228	3044.5604167	1924.5	1631	1465	24621	.	.	.	.	.	.	.	.	.	.
2011	1875.0252101	1341	970	816	17720	4207.5610973	2292	1670.2	1448	32716	.	.	.	.	.	.	.	.	.	.
2012	2027.4355828	1425	908.4	696	13792	5150.092081	2881	1694.6	1511	58052	.	.	.	.	.	.	.	.	.	.
2013	2784.5253165	1816	1028.1	723	21277	5264.6238532	3475	1744.6	1555	26092	.	.	.	.	.	.	.	.	.	.
2014	2980.2478632	1751	1059.6	803	48092	6056.7219662	6528	1820	1555	25425	5995.3	4192	2273.2	2024	39128	28.5859375	20	7	2	167
2015	2776.1131387	1956.5	1128	769	38532	6748.9140351	6721	1888.3	1559	39505	5214.9479167	3948	2408.4	2014	23096	39.540682415	28	8	1	467
2016	3986.220339	2274	1270.4	810	31200	6923.3292683	6874	2349.1	1561	30594	6429.3275862	4657	2371.6	2150	51380	29.328042328	18.5	3	1	232
2017	4789.3233083	3553	1498.6	650	26916	5189.7158879	4531	1919.6	970	76657	5210.5744681	3391	1992.2	1770	17701	32.046575342	19	3	0	398
2018	3795.3040541	3094.5	1325.3	143	17737	3806.1096207	3254.5	1162.2	751	76966	2711.3589744	2226	1714	1563	5878	51.414673046	23	3	1	6303
2019	1737.2914573	1556	487	31	5581	2063.5736926	1358	417.4	154	63797	1704.0967742	1293	839.6	788	8113	35.621076233	23	5	1	355
2020	673.74264706	458.5	53.4	13	5919	555.81763181	403	168	11	10779	356.83333333	225	66	66	1018	28.075875486	12	1	1	444

 

Based on this information, I would propose the following view thresholds for messages between 2007-2017 that are candidate for being archived:

Blogs - 700

Discussions - 1650

File Exchange - 2250

InsideJMP - 15

 

Using these numbers would result in potentially 833 messages being archived. Here's the breakdown by category:

 

 

Re: Content Archiving

Did some further exploration of the Discussions area. The Khoros data gives the total views of an article during its lifetime. The Google Analytics data gives the total views of an article during a given time span. I pulled the GA data for one year then took the intersection of the top 50 (lifetime) viewed articles and the GA results. The graphs for those can be found on JMP Live Sales under the Customer Care group.

 

There are some articles that have been viewed often in the past year but did not make the top 50 list for lifetime views. Will need to do further investigation for those.

 

The articles show in the graphs above though will be a good place to start to verify existing content is relevant.