Does your sales proposal mirror your customer’s language?

I was working with one of our VisibleThread users recently. They cited this example as one issue raised during a debrief for a lost bid; their customer talked about ‘risk abatement’ in the RFP, they talked about ‘risk mitigation’ in the proposal.

Mirroring - Obama and Cameron Now, most sales professionals are familiar with the psychology of mirroring. During a conversation, you make a gesture and the person you’re talking with makes a similar gesture. Or you start using the same idioms as the person you’re chatting with. Psychologists call this ‘mirroring’. It suggests agreement and alignment.

We analyze a lot of proposal content and we see frequent examples where proposals do not mirror the customer’s language. Examples include; over-use of your company name or inconsistent technical terminology in your proposal.

Lack of alignment damages probability of win. The government or commercial reviewer makes conscious and unconscious value judgments based on your proposal language. So if the reviewer perceives your proposal as predominantly stock boilerplate, with no attempt to mirror your customer’s language and needs, it impacts negatively.

So how do you know if you’re aligned?

Using ‘Discovery’, which shows us what concepts are at play across documents, we can identify lack of language mirroring. As an example, let’s assume you have this statement in the SOW (Statement of Work); “Assisting users in making use and taking advantage of data publication services”. The concept we need to cross check is “data publication services”. Do we talk about that in our proposal?

Let’s see how we can check.

Step 1: Upload your solicitation and response docs for analysis.

We take a number of solicitation and proposal documents, a mixture of PDF and MS Word formats, and upload them into VisibleThread.

Uploading docs for analysis

Step 2: Next Click ‘Discovery’ and analyze critical concepts.

Back to our example statement in the SOW; “Assisting users in making use and taking advantage of data publication services”. The concept we need to cross check is “data publication services”.

We extract the ‘concepts’ from the document using NLP (Natural Language Processing) techniques. You can see in the screen shot below that most of the concepts are ‘data’ related. We have 4 occurrences of ‘data publication services’, 2 occurrences of ‘climate data services’ etc. And we see the document content extracted in the bottom grid. This makes it really easy to review.

SOW showing concepts from solicitation
Step 3: So, let’s check if we’re mirroring these concepts in our proposal…

For this we look at a folder level view. Folder views are nice because they allow us compare multiple documents in one view.

The folder in this case contains 4 documents, 3 are solicitation related (including our SOW) and the 4th is our proposal. We can see the concepts arranged on the left, each column on the right hand side is a document. So you start to quickly see lack of alignment between the solicitation docs and the proposal. In this case, it’s pretty clear that we are not addressing; ‘data publication services’. Not only that, we are also missing references to ‘real-time data systems’ and ‘satellite data’. In the screen shot below, the yellow items are gaps in our proposal.

This lack of language alignment (or lack of mirroring) can be more serious. It can imply a material gap where we are missing vital content. In fact, a reviewer may view this poor alignment as making the proposal non-compliant.

Poor alignment between solicitation and proposal

Quite often we see another variation on this theme. In more complex proposals, the proposal itself may use inconsistent language. One section may not mirror another section, especially where multiple SMEs are contributors.

I came across this example recently; SME (Subject Matter Expert) 1 let’s call her “Jill” was very keen to talk about ‘onthologies’, while her colleague “Jim”, (SME 2) liked to talk ‘taxonomies’.

They were referring to the same concept in the Tech Volume. The issue here was that both terms were intermingled and therefore the reviewer will be left reeling. In this case, the anomaly was not clear until the Proposal Manager ran the above NLP based analysis on the documents in VisibleThread.


  • If you do not mirror your customer’s terminology in your sales proposal, you show lack of alignment. You may even be inadvertently non-compliant.
  • Language alignment (or the lack of it) between solicitation and response is hard to spot using manual methods.
  • With NLP techniques (available in VisibleThread), you can quickly identify gaps, and improve alignment and compliance.


If you want to see how we scan documents, check out these 3-minute demos to get a good sense.

If you want to see how you can check your documents for compliance, sign up for our no-obligation free trial.




  • Paul Spottswood
    September 11, 2014

    Good Blog,

    Any idea when version 2.11 will be released?


    • Fergal
      September 11, 2014

      Hi Paul,

      Thanks. To your question, for on-premise customers like yourself, 2.11 is tracking for early October. We’re in final QA cycles. The capability in the blog post is live already for cloud customers.

      The support guys will notify you (and all our customers) with full deployment details once available.