Anonymised project example

How a consultancy automated its lead routing

About 800 enquiries a month came in through 5 channels. A workflow in Pharen Hub gives each one a team and saves over 90 hours of searching a month.

Book a no-obligation intro call

Industry: Consulting
Services: Process design, Lead routing, CRM handoff

Over 90 hours

less searching a month, more than 2 working weeks

About €4,200

a month, at €45.00 per working hour

Same day

for most first replies, which often took 2 days before

In short

A consultancy receives about 800 enquiries a month, through its website form, by email, as referrals, on LinkedIn and from conversations. Today they all land in one list in Pharen Hub, and a workflow handles the lead routing. It checks whether the contact already exists, an agent classifies new enquiries by need, relevance and timing, a rule hands them to the team for the right region, and every weekday at 8 am the Hub shows where a follow-up is due.

Before, each enquiry took about 8 minutes of searching, over 100 hours a month in total, and every 5th enquiry got no follow-up. Today it takes about 1 minute per enquiry. That saves over 90 hours a month, about €4,200 at €45.00 per working hour, the German average for 2025. The first reply usually goes out the same day instead of often after 2 days. People still run the conversations.

With Pharen, the team has:

brought about 800 enquiries a month from 5 channels into one list
cut the search per enquiry from about 8 minutes to about 1, over 90 hours a month
saved working time worth about €4,200 a month
given every enquiry a team and a follow-up date, including the roughly 160 a month that used to be left waiting
brought the first reply forward from often 2 days to usually the same day
recognised known contacts before anyone writes to them a second time

Starting point

The consultancy had plenty of enquiries, about 800 a month, almost 40 on a normal working day. The trouble was that they came from everywhere: the website form, email, referrals, LinkedIn and personal conversations. It was often unclear what was actually new, what had already been answered and who should get in touch next.

The team did not want an aggressive sales machine. It wanted to spot good conversations sooner and follow up reliably, without forcing every enquiry straight into an overloaded CRM.

Every enquiry stayed where it arrived

A LinkedIn chat stayed on LinkedIn. A referral sat in an inbox. A website form was forwarded. That worked while a few people could keep everything in their heads.

So before anyone could reply, they had to search. Does this person already exist? Has someone already replied, and who is responsible? That meant going through the inbox, LinkedIn messages and notes, about 8 minutes per enquiry. With 800 enquiries, it came to over 100 hours a month. At €45.00, what an hour of work cost on average in Germany in 2025 according to the Federal Statistical Office, the searching alone cost about €4,800 a month.

Every 5th enquiry was left waiting

Because nobody was clearly responsible, everyone assumed someone else would reply. The first reply often took 2 days. About 160 enquiries a month, every 5th one, never got a second message after the first. Some contacts were approached twice, others not at all.

Research published in Harvard Business Review in 2011 shows what that waiting costs. Among 2,241 US companies, the first reply to a test enquiry took 42 hours on average, and 23 percent never replied. A second analysis of 1.25 million enquiries at 42 companies found that trying to make contact within an hour led to a conversation with a decision maker nearly 7 times as often as trying an hour later, and more than 60 times as often as waiting 24 hours or longer.

Implementation

Lead routing with one list for 5 channels

In Pharen Hub we set up one list for all enquiries and a workflow that fills it and routes them. Enquiries from the website form are added by the workflow itself. For forwarded emails, an agent reads each message and picks out the name, region and request. Referrals, LinkedIn messages and conversations go into the same list by hand, and they start the same process.

No contact needs a full assessment at the start. Every enquiry first needs a source, a short classification, a team and a next step.

Lead routing as a workflow in Pharen Hub in 4 lanes. At the top it adds enquiries from the website form and the email inbox to the list. In the middle it checks every new enquiry against existing contacts, an agent classifies new ones, a rule picks Team Nord, Ost, Süd or West, then it schedules the follow-up and hands the contact to the CRM. At the bottom it looks for due, still open enquiries every weekday at 8 am and flags them. — The workflow, rebuilt with sample data in Pharen Hub. At the top, the website form and the inbox add new enquiries, in the middle every new enquiry runs from the contact check to the CRM, and at the bottom the Hub looks for due enquiries every weekday at 8 am. Open the full-size view.

An enquiry moves through 5 steps

Check the contact. The workflow checks whether the person already exists. If a conversation is already running, the enquiry goes to the existing contact as a note, and nobody writes to the person a second time.
Classify the enquiry. For new contacts, an agent suggests a classification by need, relevance and timing: first call, documents first or follow up later.
Pick the team. A rule hands the enquiry to Team Nord, Ost, Süd or West, depending on the region.
Plan the follow-up. The enquiry gets a status and a date on which the team gets back in touch.
Hand over to the CRM. The contact goes to the CRM, and the enquiry is routed.

Every weekday at 8 am, the Hub looks for enquiries whose date has come and that are not yet closed, and adds a reminder to the contact.

A classification alone would not have changed anything. Enquiries were left waiting because nobody was responsible. That is why AI process automation connects the agent’s suggestion with a rule for the team and a date.

The list Anfragen in Pharen Hub with 12 sample enquiries, each with name, source, region, request, classification, team, status, follow-up date and email — The shared list with sample data. Each of the 12 enquiries has a team and a follow-up date. 3 are new, 3 have a first call planned and 2 are waiting for documents. David L.'s enquiry is closed, so the Hub no longer reminds the team about it. Open the full-size view.

People run the conversation

The workflow does not message anyone and does not turn anyone down. The classification is a suggestion that the responsible person can correct. The process is there to help the team coordinate, not to score people in ways nobody can follow. Why a short enquiry does not yet justify a judgement about the customer is the subject of the article on AI lead qualification.

Result

Over 90 hours less searching a month

Whether the contact already exists, who is responsible and when to follow up is in the list before anyone opens the enquiry. Each enquiry now needs about 1 minute to look over the classification and team, instead of about 8. Over 100 hours a month have become about 13.

That saves the team over 90 hours a month, more than 2 working weeks. At €45.00 per working hour, that is about €4,200 a month.

First reply on the same day

Today the first reply usually goes out the same day, where it often took 2 days before. The roughly 160 enquiries a month that used to be left waiting now have a team and a date. Once that date arrives, the Hub reminds the team until the enquiry is closed.

If every 6th of these 160 enquiries books, that would be about 27 additional bookings a month. That is not a promise of any particular conversion rate. Whether a conversation turns into a booking depends on the offer, the timing and the conversation itself. What the process does influence can be measured: time to first reply, enquiries without a team and overdue follow-ups.