Çağrı merkezi süreçlerinin süreç madenciliği ile incelenmesi ve maliyet analizi
Discovering call center processes using process mining and investigation of cost effect of improvements
- Tez No: 637223
- Danışmanlar: DR. ÖĞR. ÜYESİ CEMİL CEYLAN
- Tez Türü: Yüksek Lisans
- Konular: Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2020
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Endüstri Mühendisliği Ana Bilim Dalı
- Bilim Dalı: Endüstri Mühendisliği Bilim Dalı
- Sayfa Sayısı: 93
Özet
Günümüzde kullanılan yazılım ürünleri kendi görevlerini yerine getirirken bir yandan yüksek boyutta veri oluşturmaktadır. Üretilen veriler yeni uygulamalara kapı açmaktadır. Süreç Madenciliği bu veriler ile ortaya çıkmış bir alandır. Olay kayıtlarından ilgili sürecin modeli çıkarılabilmekte, çıkarılan modellerin üzerinde tekrar olay kayıtları incelenerek doğruluğu kontrol edilmekte ve problem noktaları tespit edilerek iyileştirmelerin yapılması sağlanmaktadır. Çağrı merkezleri, firmaların müşterileri ile iletişim konusunda kullandıkları öncelikli noktalardan biridir. Müşterilerin problemlerini ve isteklerini ilettikleri çağrı merkezleri, müşteri açısından önemli bir nokta olmasının yanında firmalar için de üzerinde çalışılması gereken bir alan olmuştur. Çözülmesi gereken konular ve organize edilmesi gereken büyük bir organizasyon olması nedeniyle birçok çalışma bu alanda yapılmıştır. Sesli yanıt sistemleri ise otomasyon veya yönlendirme sağlanarak kolayca çözülebilecek problemlerin çağrı merkezindeki yoğun kullanıma dahil olmadan çözülebilmesi için kullanılan teknolojidir. Çağrı merkezlerinde oluşan yoğunluğu azaltmaktadır. Bu bilgiler doğrultusunda çalışmada olay kayıtları ile elde edilen süreçler ile müşterilerin hangi durumlarda çağrı merkezi personeline yönlendirildiği incelenecektir. Bu çağrıların çağrı merkezi personeli ile görüşme yapılmasına gerek kalmadan tamamlanıp tamamlanamayacağı ve hangi düzenlemelerin gerektiği incelenecektir. Düzenlemesi yapılarak sadece sesli yanıt sistemi üzerinde sonlanması durumunda elde edilecek tasarruf hesaplanacaktır. Bir bankaya ait çağrı merkezine ulaşan bir haftalık çağrıları içeren olay kayıtları üzerinde çalışma gerçekleştirilmiştir. Çalışmada olay kayıtlarına uygun olacak şekilde Disco uygulaması ve algoritma olarak bulanık modeller kullanılmıştır. Olay kayıtlarında gerekli düzenlemeler gerçekleştirildikten sonra Disco uygulaması ile süreçler ortaya çıkarılmıştır. Yapılan çalışma doğrultusunda gerekli filtrelemeler elde edilen süreçlere uygulandıktan ve incelemeler tamamlandıktan sonra 4 adet süreç için iyileştirmelerin yapılabileceği tespit edilmiştir. Çağrı merkezi personeline yönlendirilmeden sonlanması sağlanabilecek 4 farklı süreç, 1 haftalık veri seti içerisindeki 24,897 adet vakayı etkilemektedir. Gartner raporuna dayanarak yapılan maliyet analizine göre 4 sürece ait 24,897 adet vakadan bir haftalık süreçte 126,729.85$ tasarruf sağlanabileceği görülmüştür. 52 haftalık ortalamada ise 6,537,952.2$ tasarrruf beklenmelidir. Bunun yanında çalışma kapsamında iyileştirmeler öncesi ve sonrasında incelenen 4 süreçte görülen diğer değişiklikler incelenmiştir.
Özet (Çeviri)
The software products used today, create high amount of data while performing their own tasks. Thanks to the event logs that record the behavior of the software product as well as the machine and human behavior that interacts with it, the desire to see and think about the operation of the systems has emerged. The produced data creates new opportunities for new applications. Process Mining is an area that has emerged with this data. The model of the related process can be discovered from the event logs, the event logs are checked again on the discovered models and their fitness is checked, and bottleneck points are identified and improvements are made. Process discovery has 4 quality criteria: Fitness, Precision, Generalization and Simplicity. With these four quality criteria, we can define the relationship between the data we have and the processes we obtain. Call centers are one of the priority points used by companies for communication with their customers. Call centers, where customers convey their problems and wishes, are an important point for the customer as well as an area that should be handled carefully for companies. Lots of academic research has been done in this subject because of the issues to be solved and this large organization to be organized. Interactive voice response systems, are the technology used to solve problems that can be solved easily by providing automation or guidance without having interaction with call center agent. Interactive voice response systems communicates with real people's voice or telephone dialing. In response to the voice commands or keystrokes given by the customers, a pre-prepared steps are proceeded. The customers are expected to solve their problems, convey their requests or complete them without any assistance in these scenarios, 24 hours a day, 7 days a week. It reduces the workload that occurs in call centers. In this study, we will carry out the processes discovered with the event logs will examine the situations in which the customers are directed to the call center agent. It will be examined whether these calls can be completed without the need for an interaction with the call center agents and what improvements are required to obtain it. By applying the improvement, savings and cost effect will be calculated if these processes starts and ends on the interactive voice response system. In the literature research, process mining studies and call center studies were examined. Process mining studies on many different subjects were investigated and different process analysis methods were examined. Literature studies on call center, which is one of the subjects of our study, were examined and information was obtained about the subject. Our work has been shaped with the information obtained. Studies addressing call center and process mining issues were examined. It was checked which topics they dealt with in what way. In the studies examined, it was observed that the statistical approaches or their fitness with the existing process were checked in the examination of the processes of the call centers. xx In this study, Process Mining, information technologies and banking call center issues will be discussed together. The discovery of processes based on event logs and the statistical approach to processes will be the part of the study that covers Process Mining. The part that makes our work unique is that, in addition to other studies, it will be proposed to make improvement proposals after process reviews, and examine the effects of improvement proposals in terms of cost. Thus, the effects of the improvements in the processes will be seen with a cost approach. A week of event logs of a banking call center were carried out in this study. 7-day customer data was obtained between 10.01.2020 and 16.10.2020. In order to discover processes over this data, which includes 13,426,155 events and 1,173,183 cases, data arrangements have to be made. The data is arranged to include the appropriate date format, unique case number and event number. With this process, it is prevented that the data contains any customer data. Fuzzy models as algorithm is used in Disco software. Fuzzy Mining algorithm will be used. With Fuzzy Mining algorithm, process discovery can be made based on the high number of events that do not progress within a certain order and can be directed to many different processes. After necessary organizing were made in the event logs, processes were discovered with Disco software. 102,168 different processes were discovered. In order to find out right processes for this study, the necessary filtering applied to processes. Within the scope of our study, filtering was made primarily for the calls that were started and ended successfully in the interactive voice response systems, directed to call center agent and then transferred to the interactive voice response systems. The processes performed by less than 100 cases in the 1-week period will be evaluated as exceptions and investigations will continue with 44 processes. These 44 different processes include 207,892 events and 13,160 cases. After examinations, four processes are found out for improvements. four different processes that can be ended without being directed to call center agent affect 24,897 cases in 1 week of dataset.“Limit Increase Entry Process”,“Quick Application Entry Process”,“Card Password Setting Process”and“Term Transfer Process”are four processes to be examined. Routing changes were proposed in 3 of these processes. It was observed that one of them could be improved by adding a new event. Four process improvements affect 24,897 cases in the 1-week data set examined within the scope of this study. In 3 of the four processes, the number of case steps has decreased. Likewise, in 2 of the four processes, the average call duration decreased. According to the cost analysis based on the Gartner report, the cost of completing a call on the interactive voice response systems is $ 0.45. Similarly, if a call is completed in the call center, the cost is $ 5.50. In the study, the results were obtained from the cost analysis in the Gartner report. After the improvements, the cost of each call is expected to drop from $ 5.50 to $ 0.45. 24,897 cases belonging to four processes could save $ 126,729.85 in a week. On the average of 52 weeks, it can be calculated that there will be a total of 1,294,644 cases in a year. For a year, saving of 6,537,952.2 $ should be expected. With the event logs being obtained for a longer period of time, the number of processes examined and the improvement proposals developed can be increased. The cost gain in the study was calculated from the data accessed with the Gartner report. Cost gain can be calculated based on different parameters with a different cost model to be xxi developed. Improvement scenarios can be created for different processes in accordance with this cost model. Improvement proposals have also been developed in this direction, as the work was carried out on the purpose of completing processes without being directed to the customer call center agent. In a model where different parameters affect the cost, different improvement proposals will emerge.
Benzer Tezler
- Multivariate statistical process control approach for performance evaluation in call centers
Çağrı merkezlerinde performans değerlendirmesine çok değişkenli istatistiksel süreç kontrolü yaklaşımı
TOLGA AHMET KALAYCI
Yüksek Lisans
İngilizce
2011
Endüstri ve Endüstri MühendisliğiBoğaziçi ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. ALİ RIZA KAYLAN
- Bankacılık sektöründeki operasyonel risklerin ağırlıklandırılması
Weighting operational risks in the banking sector
EZGİ BEKAR
Yüksek Lisans
Türkçe
2020
BankacılıkYıldız Teknik ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. NİHAN ÇETİN DEMİREL
- Bankacılık sektöründe dış kaynak çalışan yönetiminin iyileştirilmesinde bilgi teknolojileri kullanımına yönelik bir uygulama
An application to use information technologies to improve management of outsourced employee in the banking industry
SEREN AKBABA
Yüksek Lisans
Türkçe
2022
İşletmeİstanbul Teknik Üniversitesiİşletme Mühendisliği Ana Bilim Dalı
DOÇ. DR. NİHAN YILDIRIM
- Business process improvement: An application for call centre operations
İş süreçleri gelişimi: Çağrı merkezi operasyonları için uygulama
NESİME DENİZ EROL
Yüksek Lisans
İngilizce
2015
Endüstri ve Endüstri MühendisliğiBahçeşehir ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. AHMET BEŞKESE
- A simulation-based staff scheduling analysis for call centersin cargo industry
Kargo sektöründe çağrı merkezleri için simülasyon tabanlı personel çizelgeleme analizi
ÖMER FARUK KANDAZ
Yüksek Lisans
İngilizce
2024
Endüstri ve Endüstri Mühendisliğiİstanbul Teknik Üniversitesiİşletme Mühendisliği Ana Bilim Dalı
DOÇ. DR. TUNCAY ÖZCAN