Skip to main content
Solution

Answer before the ticket is submitted

Deflection is not about hiding the contact form. It is about being faster than it - and then measuring the result in a way your finance team would accept.

  • In-flow suggestions
  • Honest measurement
  • Gap-driven backlog

What is support ticket deflection?

Ticket deflection is the reduction in support contacts that results from customers finding answers themselves. It is achieved by putting the right answer in front of the customer at the moment they are about to ask - in search, in the product, and in the contact form itself - and it is measured by comparing support contacts per active customer against a baseline taken before the knowledge base launched.

  • Deflection happens in the support flow, not in the footer.
  • AI answers with citations, so customers trust and act on them.
  • Suggested articles appear as the customer types their ticket subject.
  • Measured against your own baseline, not a vendor benchmark.
  • Escalated and abandoned sessions reported separately from resolved.
  • Failed searches become a ranked content backlog.
The four moments

Where a ticket gets avoided

  1. 01

    In search

    The customer searches, gets a written answer with sources, and never reaches the contact page. The largest single contributor for most teams.

  2. 02

    In the product

    An embedded widget on the screen where the problem occurs - the billing page, the failed import, the empty state - answering in context.

  3. 03

    In the contact form

    Suggested articles appear as the customer types their subject line. The last, cheapest chance to deflect, and one most teams never implement.

  4. 04

    Beside the agent

    Not deflection exactly, but the same content: agents answer faster with suggested articles they can insert into a reply in one click.

In detail

Measuring it in a way that survives scrutiny

Define the number before you report it

Deflection metrics have a bad reputation because most of them are defined to flatter the tool. Ours are defined to be defensible.

  • Baseline: support contacts per active customer over the quarter before launch.
  • Deflected: a documentation session where the reader reached an answer and opened no ticket within 24 hours.
  • Escalated: the reader chose or was routed to a human. Reported, never counted as success.
  • Abandoned: the reader left mid-session. Reported separately - it usually indicates a content problem.
  • Seasonality and release spikes plotted on the same chart, so you are not fooled by a quiet month.

We would rather hand your support director a smaller number they can defend than a larger one they have to caveat in every meeting.

The loop that keeps it improving

Deflection is not a launch, it is a monthly habit. The reports are built to make that habit cheap.

  • Top failed searches, clustered by intent and ranked by volume.
  • Articles with high traffic and poor helpfulness scores - existing content that is failing.
  • Ticket topics with no corresponding article, pulled from your helpdesk categories.
  • Create a draft from any of these in one click, assigned with a due date.
  • Track whether the query volume for a closed gap actually falls afterwards.
Frequently asked

Questions people ask before they start

For a team with a reasonable existing help centre, 15-25% of first-line contacts within two quarters is a defensible target. Teams starting from nothing sometimes see more, because the baseline was so poor. Anyone quoting 60% is measuring something other than tickets.

A session counts as deflected when a reader found an answer and opened no ticket within 24 hours, compared against your own pre-launch tickets per active customer. Escalations and abandonments are reported separately and never counted as success.

Mostly in three places: search results before the customer reaches the contact form, suggested articles as they type the ticket subject, and the chatbot. The footer link to the help centre contributes far less than people expect.

Only if you make the human route hard to find. Deflection that comes from a faster answer improves satisfaction; deflection that comes from hiding the contact button does the opposite and shows up in your CSAT within weeks.

Take a baseline this month. Compare it next quarter.

Start free for 14 days. No credit card, no setup fee, and your content is yours to export at any time.

Questions first? Email sales@thedocs.in or call +91 8585953085.