I'd like to clarify the topic statement.
In many sources on the internet the terms:
ICW (in common with)
and
Triangulation
are used in the same sense and are used equally. This is wrong!
ICW is a term that is especially used by FTDNA in their chromosome browser. Many people are assuming that because three people (let's name them A - which is yourself, B and C) are in-common-with each other that also automatically means they triangulate as well.
To clarify the term triangulation. It refers to a group of three individuals (they shouldn't be closely related like 1-2 generations) that are all matching at the same locus (location on a chromosome) with overlapping matching segments between them (they can start/end sometimes earlier or later than with others in the group but they must overlap). Triangulation is the only method to identify an ancestral segment that was given down from a common ancestor that all individual in this one triangulation group are descendants from.
What the ICW is only saying is that these individuals match each other. While you can see where you are (A) are matching B and also where you match C, there is no way you (alone) can see where B and C match each other (not on FTDNA or AncestryDNA - it's possible to see this on 23andme and on GEDmatch).
There is a likelihood that indeed all three of them match at the same locus but without the information been given by either B or C to you (A) where they match each other it's impossible to tell 100% sure.
So please contact B and C and ask them for this information and don't assume a triangulation!
I do have one comment from Doris Wheeler where she has a specific example where cousins from her maternal and paternal side are put into one ICW group by FTDNA but the resulting matching segments are on different loci (as proven by analysis on GEDmatch). Dr. Tim Janzen also stated that ICW is especially bad to use in "endogamous populations such as Mennonites or Jews because such groups generally share DNA segments with most other people with that ancestry."
I further refer to the following comment in this G2G thread by Darlene (http://www.wikitree.com/g2g/201219/dna-matches-how-do-you-spread-your-sheet?show=201700#c201700): "I recommend DNAGedcom.com in addition to Genome Mate. It helps you identify triangulated matches from Gedmatch and FTDNA"
I would like to clarify that based on the data that is available for downloading at FTDNA and AncestryDNA there is only one way to do triangulation, which is by downloading the raw DNA data from both sites to a tool like GEDmatch!
No tool in the world can work on insufficient data and do triangulation, as such the statement on DNAGedcom's page about the ADSA tool is wrong and misleading: "The Autosomal DNA Segment Analyzer (ADSA) is a tool that takes your data from Family Tree DNA or GEDMATCH and constructs tables that include match and segment information as well as a visual graph of overlapping segments, juxtaposed with a customized, color-coded In-Common-With (ICW) matrix that will permit you to triangulate matching segments without having to look in multiple spreadsheets or on different web pages."
I've asked Rob Warthen from DNAGedcom to change that misleading statement and clarify it. Again, it's not a fault of the ADSA tool, it's because FTDNA (and Ancestry) don't provide enough detailed information from their screens to do triangulation!
The only DTC DNA testing company that currently provides this information is 23andme. It also has the only tool to do triangulation with up to 5 people directly on their website (the Family Inheritance: Advanced tool).
tl;dr
Hope this G2G post clarifies that:
ICW <> triangulation
and that no tool can do triangulation without detailed data (meaning matching chromosome, with overlapping start and end position between all matching individuals).